[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nick Rosbrook]

** Changed in: systemd (Ubuntu)
   Status: Confirmed => Invalid

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Invalid
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Christopher Trahey]

Clarifying what worked for me across all the servers where I need this
config:

cat 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Christopher Trahey]

Is it possible this regressed sometime between Bionic and Focal? I am
observing this in 20.04. Early in dmesg I see my interface renamed from
eth1 to eno2, then my netplan attempts to rename with a match (on
name=eno2) to my desired name ("ceph" in this case). I see syslog
message:

systemd-udevd[844]: eno2: Config file /run/systemd/network/10-netplan-
eno2.link matches device based on renamed interface name, ignoring

but after boot and login, netplan apply is succsesful.

Steps in #11 do not work, even though I verify the link file is in the
new initrd.img at usr/lib/systemd/network/10-netplan-eno2.link

What _did_ work for me, is editing this file first to refer to the eth1
name instead of eno2 (the file was written, I believe, by netplan).
After this I can change my netplan to refer to the desired name with no
set-name rule. Obvs. I can cut out some steps as I replicate this
solution to the other servers.

To summarize: write out /etc/systemd/network/10-netplan-eno2.link in a
way that refers to eth1 (instead of eno2) and sets ultimately desired
name. update-initramfs. netplan uses desired name without rename rules.
I suppose I can/should rename that file to 10-netplan-ceph.link :-)

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Jamie Lokier]

For what it's worth, I'm still seeing set-name having no affect on a
server's only interface.

Server is using Ubuntu 19.04, with netplan.io-0.96-0ubuntu4.1, AMD64
arch.  It's a bare-metal server with a single "e1000e" ethernet device.
No cloud-init installed.

Config file /etc/netplan/01-netcfg.yaml starts like this:

network:
  version: 2
  renderer: networkd
  ethernets:
enp4s0:
  match:
name: enp4s0
  set-name: eth0
  addresses:
- $IPV4
- $IPV6
   (etc.)

The interface is configured adequately in boot, but the interface name
is not changed according to "set-name eth0".  It just doesn't change,
with "netplan apply" or not.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.96-0ubuntu0.18.10.3

---
netplan.io (0.96-0ubuntu0.18.10.3) cosmic; urgency=medium

  * debian/patches/git_revert_explicit_renderer_def_ebc212a.patch: revert
commit ebc212a: make renderer values explicit at the end of each parsing
pass; it breaks "default" renderer behavior when multiple files may set
a global renderer and expect the last to take effect globally.
(LP: #1825206)
  * debian/patches/git_reorg_netdef_validation_181b583.patch: correct the
fallout from the above change: validate netdefs in a single pass at the
very end of parsing, once we know which is the applicable renderer. This
makes sure tunnels get validated correctly.

netplan.io (0.96-0ubuntu0.18.10.2) cosmic; urgency=medium

  * d/p/0001-Partially-revert-the-change-for-enabling-systemd-net.patch:
Partially revert changes to networkd jobs ordering: leave systemd-networkd
enabled in multi-user.target instead of network-online.target, as in some
minimal setups there might be no job requiring network-online.target,
whereas the natural target to reach for booting is multi-user.target.
(LP: #1821867)

netplan.io (0.96-0ubuntu0.18.10.1) cosmic; urgency=medium

  * Backport netplan.io 0.96 to 18.10.
  * debian/patches/glib_changes.patch: Patch tests to work again on older GLib.
  * debian/control: adjust Depends for cosmic / re-add nplan package.

netplan.io (0.96-0ubuntu1) disco; urgency=medium

  * New upstream release 0.96.
- Moved netplan-wpa@ services to earlier at boot (LP: #1819014)
- Restart services when unconfiguring (LP: #1811868)
- Use the .nmconnection extension for NM keyfiles (LP: #1817655)
- Fixed integration tests runner to correctly report failures
- Enforce integrity for use-routes in networkd backend.
- Ensure terminal state at end of test (LP: #1817660)
- Various small test fixes.
- Fix typos in documentation.
  * debian/control: Update Maintainer for ubuntu upload.

netplan.io (0.95-2) unstable; urgency=medium

  * Set Priority to optional (Closes: #920327).

netplan.io (0.95-1) unstable; urgency=medium

  * New upstream release.
  * Update autopkgtests from the upstream.
  * Add debian/watch following GitHub releases.
  * Add Homepage (Closes: #917233).

netplan.io (0.95) disco; urgency=medium

  * New upstream release:
- Added support for WPA Enterprise / 802.1x authentication (LP: #1739578)
- Added support for setting up IP tunnels; supporting the types: ipip,
  gretap, VTI, ISATAP (NetworkManager only), sit, gre, ipip6 and ip6ip6.
  + Fixes sit (ipv6) tunnels using Hurricane Electric (LP: #1799487)
- Add support to override networkd UseMTU setting (LP: #1807273)
- Generate output files in dependency order
- Refactored unit and integration tests, along with various cleanups.
- Add DHCP overrides to control route usage and default metric for DHCP
  routes. (LP: #1776228)
- Mitigate against bad matching on devices behind bonds then they share
  the same MAC from a physical interface. (LP: #1804861)
- Added snapcraft.yaml.

netplan.io (0.90.1) disco; urgency=medium

  * Do not assume /etc/network exists in postinst, as netbase 5.5 no longer
creates it.
  * netplan/cli/commands/ip.py: fix a flake.

netplan.io (0.90) disco; urgency=medium

  * New upstream release:
- build: fixes for building on RPM-based distros
- build: code prettiness changes (make indentation consistent)
- Fix device name-changes detection (LP: #1770082)
- Add support for IPv6 Privacy Extensions (LP: #1750392)
- Add dhcp{4,6}-overrides to control DNS, NTP, hostname updates via DHCP
  (LP: #1759014)
- Clarify MAC and MTU setting requirements (LP: #1800668)
- Various documentation fixes (LP: #1800669)
- Improve error reporting to give clearer messages and context
  (LP: #1800670)
- Skip non-physical/bond interfaces when applying renames (LP: #1802322)

 -- Mathieu Trudel-Lapierre   Fri, 26 Apr 2019
15:19:19 -0400

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.96-0ubuntu0.18.04.4

---
netplan.io (0.96-0ubuntu0.18.04.4) bionic; urgency=medium

  * debian/patches/git_revert_explicit_renderer_def_ebc212a.patch: revert
commit ebc212a: make renderer values explicit at the end of each parsing
pass; it breaks "default" renderer behavior when multiple files may set
a global renderer and expect the last to take effect globally.
(LP: #1825206)
  * debian/patches/git_reorg_netdef_validation_181b583.patch: correct the
fallout from the above change: validate netdefs in a single pass at the
very end of parsing, once we know which is the applicable renderer. This
makes sure tunnels get validated correctly.

netplan.io (0.96-0ubuntu0.18.04.3) bionic; urgency=medium

  * debian/patches/disable-networkd-tunnels-ipip-gre.patch: disable IPIP and
GRE tunnel tests; those appear to be broken because neither the kernel nor
networkd bring up the device automatically as in other releases.

netplan.io (0.96-0ubuntu0.18.04.2) bionic; urgency=medium

  * d/p/0001-Partially-revert-the-change-for-enabling-systemd-net.patch:
Partially revert changes to networkd jobs ordering: leave systemd-networkd
enabled in multi-user.target instead of network-online.target, as in some
minimal setups there might be no job requiring network-online.target,
whereas the natural target to reach for booting is multi-user.target.
(LP: #1821867)

netplan.io (0.96-0ubuntu0.18.04.1) bionic; urgency=medium

  * Backport netplan.io 0.96 to 18.04.
  * debian/patches/glib_changes.patch: Patch tests to work again on older GLib.
  * debian/control: adjust Depends for bionic / re-add nplan package.

netplan.io (0.96-0ubuntu1) disco; urgency=medium

  * New upstream release 0.96.
- Moved netplan-wpa@ services to earlier at boot (LP: #1819014)
- Restart services when unconfiguring (LP: #1811868)
- Use the .nmconnection extension for NM keyfiles (LP: #1817655)
- Fixed integration tests runner to correctly report failures
- Enforce integrity for use-routes in networkd backend.
- Ensure terminal state at end of test (LP: #1817660)
- Various small test fixes.
- Fix typos in documentation.
  * debian/control: Update Maintainer for ubuntu upload.

netplan.io (0.95-2) unstable; urgency=medium

  * Set Priority to optional (Closes: #920327).

netplan.io (0.95-1) unstable; urgency=medium

  * New upstream release.
  * Update autopkgtests from the upstream.
  * Add debian/watch following GitHub releases.
  * Add Homepage (Closes: #917233).

netplan.io (0.95) disco; urgency=medium

  * New upstream release:
- Added support for WPA Enterprise / 802.1x authentication (LP: #1739578)
- Added support for setting up IP tunnels; supporting the types: ipip,
  gretap, VTI, ISATAP (NetworkManager only), sit, gre, ipip6 and ip6ip6.
  + Fixes sit (ipv6) tunnels using Hurricane Electric (LP: #1799487)
- Add support to override networkd UseMTU setting (LP: #1807273)
- Generate output files in dependency order
- Refactored unit and integration tests, along with various cleanups.
- Add DHCP overrides to control route usage and default metric for DHCP
  routes. (LP: #1776228)
- Mitigate against bad matching on devices behind bonds then they share
  the same MAC from a physical interface. (LP: #1804861)
- Added snapcraft.yaml.

netplan.io (0.90.1) disco; urgency=medium

  * Do not assume /etc/network exists in postinst, as netbase 5.5 no longer
creates it.
  * netplan/cli/commands/ip.py: fix a flake.

netplan.io (0.90) disco; urgency=medium

  * New upstream release:
- build: fixes for building on RPM-based distros
- build: code prettiness changes (make indentation consistent)
- Fix device name-changes detection (LP: #1770082)
- Add support for IPv6 Privacy Extensions (LP: #1750392)
- Add dhcp{4,6}-overrides to control DNS, NTP, hostname updates via DHCP
  (LP: #1759014)
- Clarify MAC and MTU setting requirements (LP: #1800668)
- Various documentation fixes (LP: #1800669)
- Improve error reporting to give clearer messages and context
  (LP: #1800670)
- Skip non-physical/bond interfaces when applying renames (LP: #1802322)

netplan.io (0.40.2-1) unstable; urgency=medium

  * Upload to Debian (Closes: #882661).

netplan.io (0.40.2) cosmic; urgency=medium

  * tests/integration.py: Mark regexes with r to pacify pycodestyle's
W605.

 -- Mathieu Trudel-Lapierre   Fri, 26 Apr 2019
15:19:47 -0400

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Committed => Fix Released

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Still verification-done on bionic and cosmic; this hasn't changed with
the new revision in -proposed.

** Tags removed: verification-needed verification-needed-bionic 
verification-needed-cosmic
** Tags added: verification-done-bionic verification-done-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Brian Murray]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.04.4 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-done-bionic
** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Brian Murray]

Hello Daniel, or anyone else affected,

Accepted netplan.io into cosmic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.10.3 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-cosmic to verification-done-cosmic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-cosmic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-done-cosmic
** Tags added: verification-needed verification-needed-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

A possible SRU regression has been reported against netplan.io
0.96-0ubuntu0.18.10.2 in LP: #1825206.  This version has been rolled
back to -proposed while the investigation is ongoing.

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Released => Fix Committed

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
   

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.96-0ubuntu0.18.10.2

---
netplan.io (0.96-0ubuntu0.18.10.2) cosmic; urgency=medium

  * d/p/0001-Partially-revert-the-change-for-enabling-systemd-net.patch:
Partially revert changes to networkd jobs ordering: leave systemd-networkd
enabled in multi-user.target instead of network-online.target, as in some
minimal setups there might be no job requiring network-online.target,
whereas the natural target to reach for booting is multi-user.target.
(LP: #1821867)

netplan.io (0.96-0ubuntu0.18.10.1) cosmic; urgency=medium

  * Backport netplan.io 0.96 to 18.10.
  * debian/patches/glib_changes.patch: Patch tests to work again on older GLib.
  * debian/control: adjust Depends for cosmic / re-add nplan package.

netplan.io (0.96-0ubuntu1) disco; urgency=medium

  * New upstream release 0.96.
- Moved netplan-wpa@ services to earlier at boot (LP: #1819014)
- Restart services when unconfiguring (LP: #1811868)
- Use the .nmconnection extension for NM keyfiles (LP: #1817655)
- Fixed integration tests runner to correctly report failures
- Enforce integrity for use-routes in networkd backend.
- Ensure terminal state at end of test (LP: #1817660)
- Various small test fixes.
- Fix typos in documentation.
  * debian/control: Update Maintainer for ubuntu upload.

netplan.io (0.95-2) unstable; urgency=medium

  * Set Priority to optional (Closes: #920327).

netplan.io (0.95-1) unstable; urgency=medium

  * New upstream release.
  * Update autopkgtests from the upstream.
  * Add debian/watch following GitHub releases.
  * Add Homepage (Closes: #917233).

netplan.io (0.95) disco; urgency=medium

  * New upstream release:
- Added support for WPA Enterprise / 802.1x authentication (LP: #1739578)
- Added support for setting up IP tunnels; supporting the types: ipip,
  gretap, VTI, ISATAP (NetworkManager only), sit, gre, ipip6 and ip6ip6.
  + Fixes sit (ipv6) tunnels using Hurricane Electric (LP: #1799487)
- Add support to override networkd UseMTU setting (LP: #1807273)
- Generate output files in dependency order
- Refactored unit and integration tests, along with various cleanups.
- Add DHCP overrides to control route usage and default metric for DHCP
  routes. (LP: #1776228)
- Mitigate against bad matching on devices behind bonds then they share
  the same MAC from a physical interface. (LP: #1804861)
- Added snapcraft.yaml.

netplan.io (0.90.1) disco; urgency=medium

  * Do not assume /etc/network exists in postinst, as netbase 5.5 no longer
creates it.
  * netplan/cli/commands/ip.py: fix a flake.

netplan.io (0.90) disco; urgency=medium

  * New upstream release:
- build: fixes for building on RPM-based distros
- build: code prettiness changes (make indentation consistent)
- Fix device name-changes detection (LP: #1770082)
- Add support for IPv6 Privacy Extensions (LP: #1750392)
- Add dhcp{4,6}-overrides to control DNS, NTP, hostname updates via DHCP
  (LP: #1759014)
- Clarify MAC and MTU setting requirements (LP: #1800668)
- Various documentation fixes (LP: #1800669)
- Improve error reporting to give clearer messages and context
  (LP: #1800670)
- Skip non-physical/bond interfaces when applying renames (LP: #1802322)

 -- Mathieu Trudel-Lapierre   Thu, 28 Mar 2019
13:57:46 -0400

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

This does not need re-testing, already in bionic/cosmic, just closed
again due to changelog for this big backport.

** Tags removed: verification-needed verification-needed-bionic
** Tags added: verification-done-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
   

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.04.3 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-done-bionic
** Tags added: verification-needed verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

No need to re-test these; it's already been included in bionic/cosmic
and bug comments are a side-effect of the changelogs for the upload with
this backport.

** Tags removed: verification-needed verification-needed-bionic 
verification-needed-cosmic
** Tags added: verification-done-bionic verification-done-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.04.2 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Released => Fix Committed

** Tags removed: verification-done-bionic
** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

Hello Daniel, or anyone else affected,

Accepted netplan.io into cosmic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.10.2 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-cosmic to verification-done-cosmic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-cosmic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Released => Fix Committed

** Tags removed: verification-done verification-done-cosmic
** Tags added: verification-needed verification-needed-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Timo Aaltonen]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.04.1 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Released => Fix Committed

** Tags removed: verification-done-bionic
** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Timo Aaltonen]

no need to re-test this on bionic/cosmic

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Committed => Fix Released

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Committed => Fix Released

** Tags removed: verification-needed verification-needed-bionic 
verification-needed-cosmic
** Tags added: verification-done verification-done-bionic 
verification-done-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Timo Aaltonen]

Hello Daniel, or anyone else affected,

Accepted netplan.io into cosmic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.96-0ubuntu0.18.10.1 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-cosmic to verification-done-cosmic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-cosmic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Released => Fix Committed

** Tags removed: verification-done-cosmic
** Tags added: verification-needed verification-needed-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.40.1~18.04.3

---
netplan.io (0.40.1~18.04.3) bionic; urgency=medium

  * Fix idempotency in renaming: bond members should be exempt from rename, as
they may all share a single MAC for the bond device. (LP: #1802322)
  * tests/integration.py: add test designed to catch the above regression.

netplan.io (0.40.1~18.04.2) bionic; urgency=medium

  * Fix typo breaking rename on 'netplan apply'. (LP: #1770082)

netplan.io (0.40.1~18.04.1) bionic; urgency=medium

  * Backport netplan 0.40.1 to 18.04. (LP: #1793309)

netplan.io (0.40.1) cosmic; urgency=medium

  * tests/generate.py: use random.sample() instead of random.choices() to
better support older pythons.
  * Deal gracefully with empty files on 'netplan apply' (LP: #1795343)

netplan.io (0.40) cosmic; urgency=medium

  * New upstream release:
- networkd: route source is PreferredSource= not From=
- Improve NetworkManager error reporting on unrenderable routes.
- Don't render ipv4 dns-search unless we have an ipv4 address.
  (LP: #1786726)
- Set permissive umask on networkd .network, .link and .netdev files
  (LP: #1736965, LP: #1768560)
- Fix support for link-scope routes. (LP: #1747455)
- Update man pages for deletion of replug code.
- Spell Gratuitous ARP correctly and make it work. (LP: #1756701)
- Many typo fixes for documentation. (LP: #1783940)
- Various build system fixes.
- Fix integration tests:
  - iproute2 output changes for link-scope routes
  - fix stability of networkd igmp-resend test
  - fix manual_addresses test now that networkd lists ~. domain
- Deduplicate code for parsing interface options
- Add support for optional-addresses.

netplan.io (0.39) cosmic; urgency=medium

  * New upstream release:
- Allow link-local addresses to be configured. (LP: #1771704)
- Forces bridges with no addresses to be brought online. (LP: #1736975)

netplan.io (0.38) cosmic; urgency=medium

  * New upstream release:
- Write udev .rules files to /run/udev/rules.d to enforce interface
  renaming. (LP: #1770082)
- Don't traceback for 'netplan ip leases' when iface is not managed or
  doesn't DHCP (LP: #1768823)
- Fix duplicate "/" path separator in error messages (LP: #1771440)
- Fix incorrect terminal reset in 'netplan try' on Ctrl-C. (LP: #1768798)
- Updated doc entries: mtu, fix fwmark->mark, cleanup optional.
  (LP: #1768783)
- Added documentation validation at build.
- Added configuration example for multi-ip interfaces.
  * tests/integration.py: fix test_eth_and_bridge autopkg test harder.
  * debian/control:
- Add iproute2 to Depends.
- Add python3-netifaces to Depends, Build-Depends.

 -- Mathieu Trudel-Lapierre   Wed, 21 Nov 2018
14:42:59 -0500

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.40.2.2

---
netplan.io (0.40.2.2) cosmic; urgency=medium

  * Fix idempotency in renaming: bond members should be exempt from rename, as
they may all share a single MAC for the bond device. (LP: #1802322)
  * tests/integration.py: add test designed to catch the above regression.

netplan.io (0.40.2.1) cosmic; urgency=medium

  * Fix typo breaking rename on 'netplan apply'. (LP: #1770082)

 -- Mathieu Trudel-Lapierre   Wed, 21 Nov 2018
14:28:42 -0500

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Resetting the tags to verification-done as per the discussion above.

** Tags removed: verification-needed verification-needed-bionic 
verification-needed-cosmic
** Tags added: verification-done-bionic verification-done-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Adam Conrad]

Hello Daniel, or anyone else affected,

Accepted netplan.io into cosmic-proposed. The package will build now and
be available at https://launchpad.net/ubuntu/+source/netplan.io/0.40.2.2
in a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-cosmic to verification-done-cosmic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-cosmic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-done-cosmic
** Tags added: verification-needed verification-needed-cosmic

** Tags removed: verification-done-bionic
** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

The SRU has been rolled back while the fix for the regression is being
prepared.  I still consider the previous SRU verification valid so am
resetting the bug state but not the tags.

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Released => Fix Committed

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Released => Fix Committed

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Dmitrii Shcherbakov]

This change introduces a regression: calling netplan apply is not
idempotent for bonds, see
https://bugs.launchpad.net/netplan/+bug/1802322

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.40.2.1

---
netplan.io (0.40.2.1) cosmic; urgency=medium

  * Fix typo breaking rename on 'netplan apply'. (LP: #1770082)

 -- Mathieu Trudel-Lapierre   Mon, 22 Oct 2018
14:58:40 -0400

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

@Mathieu: Me and Nicorac have the same bug: if set-name is used, there will be 
no IP at boot. The bug 1768827 is about the same problem as this one (set-name 
not renaming interface at boot), which is resolved by now...
The strange thing is the inconsistency between the two installation types 
(subiquity and the legacy debian-installer). Please, could you try and install 
a vm with the debian-installer and try it yourself? Why does one work and the 
other doesn't? 

Created bug 1802004, IMO 17868827 should be closed, as renames are
working as expected now...

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.40.1~18.04.2

---
netplan.io (0.40.1~18.04.2) bionic; urgency=medium

  * Fix typo breaking rename on 'netplan apply'. (LP: #1770082)

netplan.io (0.40.1~18.04.1) bionic; urgency=medium

  * Backport netplan 0.40.1 to 18.04. (LP: #1793309)

netplan.io (0.40.1) cosmic; urgency=medium

  * tests/generate.py: use random.sample() instead of random.choices() to
better support older pythons.
  * Deal gracefully with empty files on 'netplan apply' (LP: #1795343)

netplan.io (0.40) cosmic; urgency=medium

  * New upstream release:
- networkd: route source is PreferredSource= not From=
- Improve NetworkManager error reporting on unrenderable routes.
- Don't render ipv4 dns-search unless we have an ipv4 address.
  (LP: #1786726)
- Set permissive umask on networkd .network, .link and .netdev files
  (LP: #1736965, LP: #1768560)
- Fix support for link-scope routes. (LP: #1747455)
- Update man pages for deletion of replug code.
- Spell Gratuitous ARP correctly and make it work. (LP: #1756701)
- Many typo fixes for documentation. (LP: #1783940)
- Various build system fixes.
- Fix integration tests:
  - iproute2 output changes for link-scope routes
  - fix stability of networkd igmp-resend test
  - fix manual_addresses test now that networkd lists ~. domain
- Deduplicate code for parsing interface options
- Add support for optional-addresses.

netplan.io (0.39) cosmic; urgency=medium

  * New upstream release:
- Allow link-local addresses to be configured. (LP: #1771704)
- Forces bridges with no addresses to be brought online. (LP: #1736975)

netplan.io (0.38) cosmic; urgency=medium

  * New upstream release:
- Write udev .rules files to /run/udev/rules.d to enforce interface
  renaming. (LP: #1770082)
- Don't traceback for 'netplan ip leases' when iface is not managed or
  doesn't DHCP (LP: #1768823)
- Fix duplicate "/" path separator in error messages (LP: #1771440)
- Fix incorrect terminal reset in 'netplan try' on Ctrl-C. (LP: #1768798)
- Updated doc entries: mtu, fix fwmark->mark, cleanup optional.
  (LP: #1768783)
- Added documentation validation at build.
- Added configuration example for multi-ip interfaces.
  * tests/integration.py: fix test_eth_and_bridge autopkg test harder.
  * debian/control:
- Add iproute2 to Depends.
- Add python3-netifaces to Depends, Build-Depends.

 -- Mathieu Trudel-Lapierre   Mon, 22 Oct 2018
15:02:30 -0400

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

@Marcos:

Please file a new bug for your own issue; include the output of
'networkctl', and report the bug number here so we can find it.

>From a quick look, I think there's no carrier detected. This could be a
driver bug.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
   

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Please use bug 1768827 to track such "no IP at boot" issues; there's
nothing that currently indicates that this is a regression, it needs
further investigation, but not here.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

@Nicorac: Tested with static IP and had same result.

ubuntu@ubuntu:~$ cat /etc/netplan/01-netcfg.yaml 
# This file describes the network interfaces available on your system
# For more information, see netplan(5).
network:
  version: 2
  renderer: networkd
  ethernets:
eth0:
  match:
  macaddress: 08:00:27:f0:97:15
  set-name: eth0
  dhcp4: no
  addresses: [ "192.168.56.2/24" ]
  gateway4: 192.168.56.1

ubuntu@ubuntu:~$ ip address show
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
   valid_lft forever preferred_lft forever
2: eth0:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
link/ether 08:00:27:f0:97:15 brd ff:ff:ff:ff:ff:ff

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

@Mathieu: I reinstalled my Virtualbox VM (using the ubuntu-18.04-server-
amd64.iso image) and tested again and the result was the same.

netplan config:
ubuntu@ubuntu:~$ cat /etc/netplan/01-netcfg.yaml 
# This file describes the network interfaces available on your system
# For more information, see netplan(5).
network:
  version: 2
  renderer: networkd
  ethernets:
eth0:
  match:
  macaddress: 08:00:27:f0:97:15
  set-name: eth0
  dhcp4: yes

Result after reboot:
ubuntu@ubuntu:~$ ip address show
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host 
   valid_lft forever preferred_lft forever
2: eth0:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
link/ether 08:00:27:f0:97:15 brd ff:ff:ff:ff:ff:ff

As you can see, the interface is successfully renamed, but the IP
address is not applied.

Do you have an idea why it works with the ubuntu-18.04-live-server-
amd64.iso image and not with ubuntu-18.04-server-amd64.iso? Perhaps it
depends on something from cloud-init package?

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

@Marcos,#80: ubuntu-18.04-server-amd64.iso is exactly the ISO I'm using
for my installations.

@Mathieu,#82: you've removed the verification-failed-bionic tag but 
0.40.1~18.04.2 is still not working on my side, both on VirtualBox VM and on 
Proxmox VM (KDE).
Interface renaming works but IP address/netmask doesn't.
I'm using ubuntu-18.04-server-amd64.iso to install both of them, and they both 
have the same issue.

PS: I actually have a production server VM witn netplan-0.40.1~18.04.2 and, 
after reboot, I need to connect to it (thanks to Proxmox HTTP-VNC console) and 
manually execute "netplan apply" to bring the network up.
This is unaccetable so I'll soon remove netplan and switch back to the rock 
solid ifupdown for such a simple newtork configuration like mine.

Sorry to say that netplan brought me nothing but issues...

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Yes, that's because it doesn't appear like the issues you have are a
regression caused by the updates, and we don't know what's wrong. You
have a bug open for your issue, and we still need to figure out what
doesn't work there.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Verification-done on bionic with netplan.io 0.40.1~18.04.2:

Using the following config:

network:
version: 2
ethernets:
maas0:
addresses:
- 10.3.21.29/20
gateway4: 10.3.16.1
match:
macaddress: 52:54:00:4d:3e:84
mtu: 1500
set-name: maas0

The interface renaming is correctly done at boot time; then changing
set-name to 'maas1', bringing down the interface with 'ip link set dev
maas0 down', and running 'netplan apply'; the interface is correctly
renamed:


ubuntu@nice-baboon:~$ sudo vi /etc/netplan/50-cloud-init.yaml 
ubuntu@nice-baboon:~$ sudo netplan apply
ubuntu@nice-baboon:~$ ip addr
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host 
   valid_lft forever preferred_lft forever
2: maas0:  mtu 1500 qdisc fq_codel state UP 
group default qlen 1000
link/ether 52:54:00:4d:3e:84 brd ff:ff:ff:ff:ff:ff
inet 10.3.21.29/20 brd 10.3.31.255 scope global maas0
   valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe4d:3e84/64 scope link 
   valid_lft forever preferred_lft forever
ubuntu@nice-baboon:~$ 
ubuntu@nice-baboon:~$ sudo vi /etc/netplan/50-cloud-init.yaml 
ubuntu@nice-baboon:~$ ip link set dev maas0 down
RTNETLINK answers: Operation not permitted
ubuntu@nice-baboon:~$ sudo !!
sudo ip link set dev maas0 down
ubuntu@nice-baboon:~$ ip addr
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host 
   valid_lft forever preferred_lft forever
2: maas0:  mtu 1500 qdisc fq_codel state DOWN group 
default qlen 1000
link/ether 52:54:00:4d:3e:84 brd ff:ff:ff:ff:ff:ff
inet 10.3.21.29/20 brd 10.3.31.255 scope global maas0
   valid_lft forever preferred_lft forever
ubuntu@nice-baboon:~$ sudo netplan apply
ubuntu@nice-baboon:~$ ip addr
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host 
   valid_lft forever preferred_lft forever
2: maas1:  mtu 1500 qdisc fq_codel state UP 
group default qlen 1000
link/ether 52:54:00:4d:3e:84 brd ff:ff:ff:ff:ff:ff
inet 10.3.21.29/20 brd 10.3.31.255 scope global maas1
   valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe4d:3e84/64 scope link 
   valid_lft forever preferred_lft forever


** Tags removed: verification-failed-bionic verification-needed
** Tags added: verification-done-bionic

** Tags removed: verification-done-xenial verification-needed-cosmic
** Tags added: verification-needed-xenial

** Tags removed: verification-needed-xenial

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Verification-done for cosmic using netplan.io 0.40.2.1:

I have verified that renames happen correctly with the config provided
in test cases and the config I have shown for the testing on bionic.

In all cases, set-name is applied as the new name for the interface, as
expected. On reboot, this is sufficient for the interface to be renamed.
After boot, the interface must first be brought down before it can be
renamed (this is expected behavior, renaming does not work on a busy
interface).

** Tags added: verification-done-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
  - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Description changed:

  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.
  
  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0
- 
+ - Bring down interface : 'ip link set dev ens3 down'
  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.
  
  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.
  
  === systemd issue ===
  
  Renaming devices doesn't seem to work.
  
  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:
  
  [Match]
  MACAddress=52:54:00:c1:c9:bb
  
  [Link]
  Name=myiface3
  
  I expect this to cause the device with that MAC address to be renamed to
  myiface3. However, when I reboot, I instead see:
  
  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff
  
  The device is not renamed.
  
  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.
  
  The renaming does work if I boot with net.ifnames=0, and oddly, it also
  works if I unbind the device and rebind it as netplan apply does. No
  setting of NamePolicy seems to help.
  
  === Original Bug ==
  
  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.
  
  Say I take this 50-cloud-init.yaml file:
  
  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3
  
  Say I change set-name to 'myiface3' and reboot. I expect that the device
  will be called myiface3 and brought up fine with dhcp. However, instead
  I see:
  
  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  
  The name has not been changed, and the device has not been brought up.
  
  If I run netplan apply however, I see the following:
  
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever
  
  So names are successfully changed with netplan apply.
  
  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.
  
  This breaks some forms of migration in certain cloud environments.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.40.2.1

---
netplan.io (0.40.2.1) cosmic; urgency=medium

  * Fix typo breaking rename on 'netplan apply'. (LP: #1770082)

 -- Mathieu Trudel-Lapierre   Mon, 22 Oct 2018
14:58:40 -0400

** Changed in: netplan.io (Ubuntu)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
    

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Tested just now... It is working for the subiquity installer version
(ubuntu-18.04-live-server-amd64.iso) but it does not apply ip address on
boot if installed with the debian-intaller version (ubuntu-18.04-server-
amd64.iso).

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Committed
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Still not working on my side (plain vanilla Ubuntu Server 18.04, fully
updated, no cloud).

The renaming regression of 0.40.1~18.04.1 has been fixed, but it's still
not applying IP address just after boot.

A manual "netplan apply" correctly configures the network interface.

-
root@test:~# dpkg -s netplan.io | grep Version
Version: 0.40.1~18.04.2
-

===
=== Just after boot ===
===

root@test:~# ifconfig -a
eth_lan: flags=4098  mtu 1500
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 6  bytes 318 (318.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 6  bytes 318 (318.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

-
root@test:~# cat /etc/netplan/01-netcfg.yaml
# This file describes the network interfaces available on your system
# For more information, see netplan(5).
network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_lan
  addresses: [ 192.168.1.10/24 ]
  gateway4: 192.168.1.1
  nameservers:
  addresses: [ 8.8.4.4 ]

-
root@test:~# cat /run/systemd/network/10-netplan-id0.link
[Match]
MACAddress=08:00:27:6b:d8:91

[Link]
Name=eth_lan
WakeOnLan=off

-
root@test:~# cat /run/systemd/network/10-netplan-id0.network
[Match]
MACAddress=08:00:27:6b:d8:91
Name=eth_lan

[Network]
LinkLocalAddressing=ipv6
Address=192.168.1.10/24
Gateway=192.168.1.1
DNS=8.8.4.4

-


=== After manual "netplan apply" ===


root@test:~# ifconfig -a
eth_lan: flags=4163  mtu 1500
inet 192.168.1.10  netmask 255.255.255.0  broadcast 192.168.1.255
inet6 fe80::a00:27ff:fe6b:d891  prefixlen 64  scopeid 0x20
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 277  bytes 51263 (51.2 KB)
RX errors 0  dropped 128  overruns 0  frame 0
TX packets 13  bytes 946 (946.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 20  bytes 1470 (1.4 KB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 20  bytes 1470 (1.4 KB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

-
root@test:~# cat /etc/netplan/01-netcfg.yaml
# This file describes the network interfaces available on your system
# For more information, see netplan(5).
network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_lan
  addresses: [ 192.168.1.10/24 ]
  gateway4: 192.168.1.1
  nameservers:
  addresses: [ 8.8.4.4 ]

-
root@test:~# cat /run/systemd/network/10-netplan-id0.link
[Match]
MACAddress=08:00:27:6b:d8:91

[Link]
Name=eth_lan
WakeOnLan=off

-
root@test:~# cat /run/systemd/network/10-netplan-id0.network
[Match]
MACAddress=08:00:27:6b:d8:91
Name=eth_lan

[Network]
LinkLocalAddressing=ipv6
Address=192.168.1.10/24
Gateway=192.168.1.1
DNS=8.8.4.4

-

Please let me know if you need further info...


** Tags removed: verification-needed-bionic
** Tags added: verification-failed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Committed
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Brian Murray]

Hello Daniel, or anyone else affected,

Accepted netplan.io into cosmic-proposed. The package will build now and
be available at https://launchpad.net/ubuntu/+source/netplan.io/0.40.2.1
in a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-cosmic to verification-done-cosmic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-cosmic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Cosmic)
   Status: Fix Released => Fix Committed

** Tags added: verification-needed-cosmic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Committed
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed
Status in netplan.io source package in Cosmic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Brian Murray]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.40.1~18.04.2 in a few
hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-failed-bionic
** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

I found time to run more tests and discovered that my patches were
wrong... And the fix was actually even simpler.

root@netplan:~# cat fix-bug-1770082.diff
Index: b/netplan/cli/commands/apply.py
===
--- a/netplan/cli/commands/apply.py 2018-10-04 16:04:57.0 +
+++ b/netplan/cli/commands/apply.py 2018-10-15 18:37:06.975347468 +
@@ -140,9 +140,9 @@
 driver = match.get('driver')
 mac = match.get('macaddress')
 if driver:
-matches['by-driver'][driver] = phy
+matches['by-driver'][driver] = newname
 if mac:
-matches['by-mac'][mac] = phy
+matches['by-mac'][mac] = newname
 
 # /sys/class/net/ens3/device -> ../../../virtio0
 # /sys/class/net/ens3/device/driver -> 
../../../../bus/virtio/drivers/virtio_net

Just changed two words...
I created a pull request on github 
(https://github.com/CanonicalLtd/netplan/pull/50).

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

IMHO if a "match" option is present (especially with a MAC condition),
it should always override the ID.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Actually, sent the wrong patch... this is the right one:

root@netplan:~# cat fix-bug-1770082.diff
--- a/netplan/cli/commands/apply.py
+++ b/netplan/cli/commands/apply.py
@@ -170,15 +170,15 @@
 link = netifaces.ifaddresses(interface)[netifaces.AF_LINK][0]
 macaddress = link.get('addr')
 if driver_name in matches['by-driver']:
- new_name = matches['by-driver'][driver_name]
- logging.debug(new_name)
+ current_name = matches['by-driver'][driver_name]
+ logging.debug(current_name)
 logging.debug(interface)
- if new_name != interface:
- changes.update({interface: {'name': new_name}})
+ if current_name != newname:
+ changes.update({interface: {'name': newname}})
 if macaddress in matches['by-mac']:
- new_name = matches['by-mac'][macaddress]
- if new_name != interface:
- changes.update({interface: {'name': new_name}})
+ current_name = matches['by-mac'][macaddress]
+ if current_name != newname:
+ changes.update({interface: {'name': newname}})

 logging.debug(changes)
 return changes

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

After some debugging, I could make it work... I'm not a programmer, so
it might be wrong, but it's working:

root@netplan:~# cat fix-bug-1770082.diff 
--- a/netplan/cli/commands/apply.py
+++ b/netplan/cli/commands/apply.py
@@ -170,15 +170,15 @@
 link = netifaces.ifaddresses(interface)[netifaces.AF_LINK][0]
 macaddress = link.get('addr')
 if driver_name in matches['by-driver']:
-new_name = matches['by-driver'][driver_name]
-logging.debug(new_name)
+current_name = matches['by-driver'][driver_name]
+logging.debug(current_name)
 logging.debug(interface)
-if new_name != interface:
-changes.update({interface: {'name': new_name}})
+if current_name != interface:
+changes.update({interface: {'name': newname}})
 if macaddress in matches['by-mac']:
-new_name = matches['by-mac'][macaddress]
-if new_name != interface:
-changes.update({interface: {'name': new_name}})
+current_name = matches['by-mac'][macaddress]
+if current_name != interface:
+changes.update({interface: {'name': newname}})
 
 logging.debug(changes)
 return changes


After applying the patch, it's working both on reboot and on netplan apply, 
with the interface down, of course.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Ok, I also re-read Daniel's comment and I see that he mentioned it
working on the previous upload. This really seems like a regression -
for the time being let's mark this verification-failed-bionic so that no
one accidentally releases it to -updates.

** Tags removed: verification-done-bionic
** Tags added: verification-failed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
   

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Just confirmed the regression... It's working as expected on boot, but is 
renaming to the "ID" key, instead of the "set-name" when live.
E.g. if we have:
ens0p3:
match:
macaddress: 01:02:03:04:05:06
set-name: eth0
It will be renamed to ens0p3 on netplan apply.
eth0:
match:
macaddress: 01:02:03:04:05:06
set-name: ens0p3
It will be renamed to eth0...

So basically, on boot it picks up "set-name", and afterwards it picks
the ID. If you keep both equal, no problem, but if have different "ID"
and "set-name"...

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

So the bug here is about renaming interfaces on boot and this seems to
be resolved by the new upload. Does this regress previous behavior? Did
renaming interfaces on a running system work in the previous version and
this upload introduces a regression? If not, we should fill in a
separate bug and try getting it fixed with the next upload.

Will be waiting with releasing this package until it's clear what's the
situation here.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

Hi,

This works on reboot but not so well after the system boots.

Say I have the following file:

network:
version: 2
ethernets:
ens7:
dhcp4: true
match:
macaddress: 52:54:00:b4:02:6e
set-name: myif1
optional: true

If I reboot, everything is fine:

ubuntu@netplan2:~$ ip l
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens3:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:d2:9b:b5 brd ff:ff:ff:ff:ff:ff
3: myif1:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:b4:02:6e brd ff:ff:ff:ff:ff:ff


But if I now change set-name to myif0, I can't get it to change the name 'live'.

If I apply with the interface 'up', nothing happens, which is fine:

ubuntu@netplan2:~$ sudo netplan apply
ubuntu@netplan2:~$ ip l
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens3:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:d2:9b:b5 brd ff:ff:ff:ff:ff:ff
3: myif1:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:b4:02:6e brd ff:ff:ff:ff:ff:ff

But if I bring it down and run netplan apply, the interface is renamed
to ens7, not myif0, and it stays down.

ubuntu@netplan2:~$ sudo ip l set dev myif1 down
ubuntu@netplan2:~$ sudo netplan apply
ubuntu@netplan2:~$ ip l
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens3:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:d2:9b:b5 brd ff:ff:ff:ff:ff:ff
3: ens7:  mtu 1500 qdisc fq_codel state DOWN mode DEFAULT 
group default qlen 1000
link/ether 52:54:00:b4:02:6e brd ff:ff:ff:ff:ff:ff

If I downgrade to 0.36.3, the interface is correctly renamed and brought
up:

9: myif0:  mtu 1500 qdisc fq_codel state UP 
mode DEFAULT group default qlen 1000
link/ether 52:54:00:b4:02:6e brd ff:ff:ff:ff:ff:ff

So not an unqualified success here.

Regards,
Daniel

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

Resetting to verification-done for bionic; using netplan.io
0.40.1~18.04.1:

I have run more tests and I can verify that renames at boot are behaving
correctly with netplan -- netplan follows exactly the configuration
provided on the system:

network:
version: 2
ethernets:
ens6:
match:
macaddress: 52:54:00:ab:37:d8
mtu: 1500
addresses:
- 10.3.21.28/20
gateway4: 10.3.16.1
nameservers:
addresses:
- 10.3.21.25
search:
- cloud.cyphermox.net
set-name: myiface11

This leads to my system having a single interface "myiface11":

ubuntu@new-man:~$ ip link
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: myiface11:  mtu 1500 qdisc fq_codel state 
UP mode DEFAULT group default qlen 1000
link/ether 52:54:00:ab:37:d8 brd ff:ff:ff:ff:ff:ff

I have removed the duplicate tag from bug 1768827 so we can keep
investigating and figure out why this isn't working for Nicorac; it
looks like there's something competing with netplan somewhere, either
from cloud-init or something else.


** Tags removed: verification-failed-bionic
** Tags added: verification-done-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Wrong cut:

---
# dpkg -s netplan.io | grep Version
Version: 0.40.1~18.04.1
---

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Sorry, still not working here. Same behavior as my previous post.
Maybe it's due to VirtualBox? Haven't tried on my production KDE (Proxmox) 
yet...

---
# dpkg -s netplan.io | grep Version
Version: 0.40~18.04.1
---


** Tags removed: verification-needed-bionic
** Tags added: verification-failed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.40.1~18.04.1 in a few
hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Tags removed: verification-failed-bionic
** Tags added: verification-needed verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Sorry for being such a buzzkill, but it's still not working on my side
:(

=
Reinstalled 18.04.1 server x64 from scratch in a clean VirtualBox VM.
=

# lsb_release -a
Distributor ID: Ubuntu
Description:Ubuntu 18.04.1 LTS
Release:18.04
Codename:   bionic

=
Enabled -proposed repo and updated netplan.io
=

# dpkg -s netplan.io | grep Version
Version: 0.40~18.04.1


Set a very basic network configuration


# cat /etc/netplan/01-netcfg.yaml
network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_lan
  addresses: [ 192.168.1.10/24 ]
  gateway4: 192.168.1.1
  nameservers:
  addresses: [ 8.8.4.4 ]


After reboot the interface is renamed
but IP address is not set 

# ifconfig -a
eth_lan: flags=4098  mtu 1500
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
...

===
After manually running "netplan apply" (after boot)
things go even worst: network interface is renamed
again to "id0"
===
# netplan apply
# ifconfig -a
id0: flags=4098  mtu 1500
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
...


** Tags removed: verification-needed-bionic
** Tags added: verification-failed-bionic

** Tags removed: verification-needed

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/netplan.io/0.40~18.04.1 in a few
hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.  Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance for helping!

N.B. The updated package will be released to -updates after the bug(s)
fixed by this package have been verified and the package has been in
-proposed for a minimum of 7 days.

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Released => Fix Committed

** Tags removed: verification-done-bionic
** Tags added: verification-needed verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

I've tried some workarounds:
- kernel parameter "net.ifnames=0" on kernel commandline: NOT WORKING
- ebtables: NOT WORKING
- add "@reboot /usr/sbin/netplan apply" to root cron: NOT WORKING

The following works on my side:

* create a new unit file: /etc/systemd/system/netplan-workaround.service

[Unit]
After=syslog.service

[Service]
ExecStart=/usr/sbin/netplan apply

[Install]
WantedBy=default.target


* enable and start it
# systemctl enable netplan-workaround.service
# systemctl start netplan-workaround.service

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

@Marcos: ebtables is ininfluent on my side.
It is installed with server version and the bug persists both with and without 
it.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Just tried with a fresh installed Ubuntu SERVER 18.04.1.
The interface now got renamed but IP address is not applied, both static and 
DHCP.

If I run "netplan apply" just after reboot+login, network interface is
configured correctly (again, both with static/DHCP address).

--
Just after reboot + login
--
# ifconfig -a
eth_lan: flags=4098  mtu 1500
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 14  bytes 910 (910.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 14  bytes 910 (910.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

--
After netplan apply
--
# netplan apply
# ifconfig -a
eth_lan: flags=4163  mtu 1500
inet 10.0.3.200  netmask 255.255.0.0  broadcast 10.0.255.255
inet6 fe80::a00:27ff:fe6b:d891  prefixlen 64  scopeid 0x20
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 28739  bytes 2625648 (2.6 MB)
RX errors 0  dropped 3379  overruns 0  frame 0
TX packets 186  bytes 19916 (19.9 KB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 197  bytes 14182 (14.1 KB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 197  bytes 14182 (14.1 KB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

--
netplan config
--
# cat /etc/netplan/01-netcfg.yaml
# This file describes the network interfaces available on your system
# For more information, see netplan(5).
network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_lan
#  dhcp4: true
  addresses: [ 10.0.3.200/16 ]
  gateway4: 10.0.0.1
  nameservers:
  addresses:
  - 8.8.4.4

I'm going to remove all of this netplan stuff and get back to old
working /ets/network/interface...

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Nicorac, I misread your message, what I meant to say is that installing
ebtables makes netplan to work properly.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Yes... I tried multiple times on VirtualBox. Installed from mini.iso
without selecting anything, then edited /etc/netplan/01-netcfs.yaml,
added match macaddress and set-name statements, rebooted and no luck.
Then installed ebtables and rebooted and it worked, then uninstalled
ebtables and it stopped working. Note that for me to discover what
package from ubuntu-server metapackage was fixing netplan I had to dig a
lot.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Do you mean that removing ebtables (if possible) let rename work
properly?

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

Can't edit, so I'm writing another message. I was wrong about ubuntu-
desktop, it does not install ebtables by default, but server does.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Marcos]

This problem still exists in a fresh (minimal) installation of Ubuntu
18.04. To reproduce, just install from mini.iso without selecting any
packages at tasksel. I've discovered that if you install ebtables it
works as expected. That's why it works if you select  an Desktop or
Server option at tasksel, because both end up installing ebtables (idk
why, but whatever).

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

** Changed in: netplan
   Status: Confirmed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Fix Released
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Since the new issue is tracked in a separate bug and currently only
reproducible by the reporter, I am releasing the current update as is.
In case this indeed turns out to be a regression, let's consider fixing
it with a subsequent upload.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package nplan - 0.32~16.04.6

---
nplan (0.32~16.04.6) xenial; urgency=medium

  [ Mathieu Trudel-Lapierre ]
  * tests/integration.py: Fix autopkgtests involving bonds/bridges to do proper
cleanup every time, so later tests don't unnecessarily wait for an
interface not configured to be up. (LP: #1775097)

  [ Daniel Axtens ]
  * Generate udev rules files to rename devices (LP: #1770082)
Due to a systemd issue[1], using link files to rename interfaces
doesn't work as expected. Link files will not rename an interface
if it was already renamed, and interfaces are renamed in initrd, so
set-name will often not work as expected when rebooting. However,
rules files will cause a renaming, even if the interface has been
renamed in initrd.

 -- Mathieu Trudel-Lapierre   Tue, 03 Jul 2018
12:55:11 -0400

** Changed in: nplan (Ubuntu Xenial)
   Status: Fix Committed => Fix Released

** Changed in: netplan.io (Ubuntu Bionic)
   Status: Fix Committed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.36.3

---
netplan.io (0.36.3) bionic; urgency=medium

  * Generate udev rules files to rename devices (LP: #1770082)
Due to a systemd issue[1], using link files to rename interfaces
doesn't work as expected. Link files will not rename an interface
if it was already renamed, and interfaces are renamed in initrd, so
set-name will often not work as expected when rebooting. However,
rules files will cause a renaming, even if the interface has been
renamed in initrd.
  * tests/integration.py: fix test_eth_and_bridge autopkg test harder.
  * tests/integration.py: fix test_mix_bridge_on_bond autopkgtest too.

 -- Mathieu Trudel-Lapierre   Thu, 05 Jul 2018
16:54:52 -0400

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Released
Status in netplan.io source package in Bionic:
  Fix Released

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

The behavior is not reproducible, maybe a race condition?
I'm actually not getting IP addres for static interface too :-(

I agree, let's continue there:
https://bugs.launchpad.net/netplan/+bug/1768827

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

I'd be a bit surprised if there really was a regression due to the
change, we're more likely seeing something going down due to the DHCP
server being slow to respond, possibly because it already responded
before?

What does the 'networkctl' command show immediately after boot?

Let's continue debugging this on your own bug (1768827).

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Sorry for the confusion.

Current version (0.36.2):
- both devices not renamed after boot (have original enp.. names)
- static and dynamic IPs correctly assigned

Proposed version (0.36.3):
- both devices correctly renamed after boot:
  :91 gets renamed eth_static
  :f5 gets renamed to eth_dhcp
- only eth_static got the configured static address, eth_dhcp not configured 
(need to call "ifconfig -a" to show it)

No VB Additions, just plain Ubuntu Server 18.04 install followed by "apt
update && apt dist-upgrade".

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  

Re: [Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

Hi,

I'm a bit confused then.

This is the netplan config you mentioned:

network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_static
  addresses: [ 1.2.3.4/16 ]
  gateway4: 5.6.7.8
id1:
  match:
macaddress: 08:00:27:23:68:f5
  set-name: eth_dhcp
  dhcp4: true


With the current version (not the -proposed version), my understanding
is that both devices keep their original names, neither device comes
up at all, and neither device will get an IP, dynamic or static. But
if I understand your comment correctly, you're saying that one of the
devices (the one with mac :f5) comes up and gets a DHCP IP? Is that
right? And if so, what is the device called - eth_dhcp or something
else?

Do you have virtualbox guest additions? I wonder if that's doing
something funky. And just checking you don't have ifupdown installed?

Regards,
Daniel

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

@Daniel: no, it's a regression to me because set-name never worked and
now DHCP stopped working too.

But I'm not testing the same environment as you, I'm on VirtualBox and this is 
my original bug report (linked to this one):
https://bugs.launchpad.net/netplan/+bug/1768827

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

Hi,

I have confirmed that the Bionic package works on the cloud service
where the bug was originally observed. With this package, the affected
migrations now succeed.

Regards,
Daniel

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

Hi,

I can verify correct behaviour with renaming, static IPs and DHCP with
KVM. I'm about to check some cloud services as well.

Nicorac, are you claiming this is a regression or just an incomplete
fix?

If it's just an incomplete fix, I think we should spin that out into
another bug and fix it after this SRU goes through.

Regards,
Daniel

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
   

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

** Tags removed: verification-failed-bionic
** Tags added: verification-done-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Partially working here on fully updated Ubuntu 18.04 Server on
VirtualBox :(

I've enabled -proposed repo and updated both nplan/bionic-proposed and
netplan.ion/bionic-proposed packages.

Now interfaces got renamed, but IPs are not assigned at boot (both DHCP and 
static ones).
Calling "netplan generate && netplan apply" after boot only sets the static IPs 
while DHCP still remains unset.


# dpkg -s netplan.io
Package: netplan.io
Status: install ok installed
Priority: important
Section: net
Architecture: amd64
Version: 0.36.3

# cat /etc/netplan/01-netcfg.yaml
network:
  version: 2
  renderer: networkd
  ethernets:
id0:
  match:
macaddress: 08:00:27:6b:d8:91
  set-name: eth_static
  addresses: [ 1.2.3.4/16 ]
  gateway4: 5.6.7.8
id1:
  match:
macaddress: 08:00:27:23:68:f5
  set-name: eth_dhcp
  dhcp4: true

*** (just after boot) ***
# ifconfig -a
eth_dhcp: flags=4098  mtu 1500
ether 08:00:27:23:68:f5  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

eth_static: flags=4098  mtu 1500
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 2  bytes 78 (78.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 2  bytes 78 (78.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0


*** (after "netplan apply") ***
eth_dhcp: flags=4098  mtu 1500
ether 08:00:27:23:68:f5  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

eth_static: flags=4098  mtu 1500
inet 1.2.3.4  netmask 255.255.0.0  broadcast 1.2.255.255
inet6 fe80::a00:27ff:fe6b:d891  prefixlen 64  scopeid 0x20
ether 08:00:27:6b:d8:91  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1000  (Local Loopback)
RX packets 2  bytes 78 (78.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 2  bytes 78 (78.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0


** Tags removed: verification-needed-bionic
** Tags added: verification-failed-bionic

** Tags removed: verification-needed

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Steve Langasek]

Hello Daniel, or anyone else affected,

Accepted netplan.io into bionic-proposed. The package will build now and
be available at https://launchpad.net/ubuntu/+source/netplan.io/0.36.3
in a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-bionic to verification-done-bionic. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-bionic. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance!

** Changed in: netplan.io (Ubuntu Bionic)
   Status: New => Fix Committed

** Tags added: verification-needed-bionic

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed
Status in netplan.io source package in Bionic:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Daniel Axtens]

Hi Łukasz,

I couldn't find the package in -proposed, but I was able to download it
from the link.

I verified that a device was not renamed with 0.32~16.04.5 but was
correctly renamed with 0.32~16.04.6, so for Xenial, verification
succeeds.

Regards,
Daniel

** Tags removed: verification-needed-xenial
** Tags added: verification-done-xenial

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Hello Daniel, or anyone else affected,

Accepted nplan into xenial-proposed. The package will build now and be
available at https://launchpad.net/ubuntu/+source/nplan/0.32~16.04.6 in
a few hours, and then in the -proposed repository.

Please help us by testing this new package.  See
https://wiki.ubuntu.com/Testing/EnableProposed for documentation on how
to enable and use -proposed.Your feedback will aid us getting this
update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug,
mentioning the version of the package you tested and change the tag from
verification-needed-xenial to verification-done-xenial. If it does not
fix the bug for you, please add a comment stating that, and change the
tag to verification-failed-xenial. In either case, without details of
your testing we will not be able to proceed.

Further information regarding the verification process can be found at
https://wiki.ubuntu.com/QATeam/PerformingSRUVerification .  Thank you in
advance!

** Changed in: nplan (Ubuntu Xenial)
   Status: New => Fix Committed

** Tags added: verification-needed verification-needed-xenial

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed
Status in nplan source package in Xenial:
  Fix Committed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test case]
  - Write a new netplan YAML (adjusting for current system as necessary):
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: myif0

  - Run 'netplan apply'
  - Verify that the device is correctly renamed to 'myif0'.
  - Reboot.
  - Make sure the device is correctly renamed to 'myif0'.

  [Regression potential]
  Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.

  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Description changed:

+ [Impact]
+ Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.
+ 
+ [Test case]
+ - Write a new netplan YAML (adjusting for current system as necessary):
+ network:
+ version: 2
+ ethernets:
+ ens3:
+ dhcp4: true
+ match:
+ macaddress: 52:54:00:de:bd:f6
+ set-name: myif0
+ 
+ - Run 'netplan apply'
+ - Verify that the device is correctly renamed to 'myif0'.
+ - Reboot.
+ - Make sure the device is correctly renamed to 'myif0'.
+ 
+ [Regression potential]
+ Changes in rename logic to add udev rules may otherwise impact applying 
different settings to the network interfaces. Changes in settings on network 
interfaces, missing parameters (especially on bonds, bridges) should be 
investigated as potential regressions. Other failures to apply network settings 
might also happen if there's a race between applying renames via the udev 
rules, and using the new names to apply configuration changes to the interfaces.
+ 
  === systemd issue ===
  
  Renaming devices doesn't seem to work.
  
  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:
  
  [Match]
  MACAddress=52:54:00:c1:c9:bb
  
  [Link]
  Name=myiface3
  
  I expect this to cause the device with that MAC address to be renamed to
  myiface3. However, when I reboot, I instead see:
  
  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff
  
  The device is not renamed.
  
  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.
  
  The renaming does work if I boot with net.ifnames=0, and oddly, it also
  works if I unbind the device and rebind it as netplan apply does. No
  setting of NamePolicy seems to help.
  
  === Original Bug ==
  
  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.
  
  Say I take this 50-cloud-init.yaml file:
  
  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3
  
  Say I change set-name to 'myiface3' and reboot. I expect that the device
  will be called myiface3 and brought up fine with dhcp. However, instead
  I see:
  
  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  
  The name has not been changed, and the device has not been brought up.
  
  If I run netplan apply however, I see the following:
  
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever
  
  So names are successfully changed with netplan apply.
  
  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.
  
  This breaks some forms of migration in certain cloud environments.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  [Impact]
  Systems relying on renaming network interfaces at boot and when 'netplan 
apply' is run.

  [Test 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Łukasz Zemczak]

Could this bug be updated with SRU relevant information? Also, it would
also be nice if we had the same changes prepared for netplan.io in
bionic.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Nicorac]

Thanks for the update.

Referring to this duplicate bug 
(https://bugs.launchpad.net/netplan/+bug/1768827), is it going to be backported 
to 18.04?
Or is there a way to install 0.38 on Ubuntu Server 18.04?

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Also affects: nplan (Ubuntu)
   Importance: Undecided
   Status: New

** Changed in: nplan (Ubuntu)
   Status: New => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in nplan package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Launchpad Bug Tracker]

This bug was fixed in the package netplan.io - 0.38

---
netplan.io (0.38) cosmic; urgency=medium

  * New upstream release:
- Write udev .rules files to /run/udev/rules.d to enforce interface
  renaming. (LP: #1770082)
- Don't traceback for 'netplan ip leases' when iface is not managed or
  doesn't DHCP (LP: #1768823)
- Fix duplicate "/" path separator in error messages (LP: #1771440)
- Fix incorrect terminal reset in 'netplan try' on Ctrl-C. (LP: #1768798)
- Updated doc entries: mtu, fix fwmark->mark, cleanup optional.
  (LP: #1768783)
- Added documentation validation at build.
- Added configuration example for multi-ip interfaces.
  * tests/integration.py: fix test_eth_and_bridge autopkg test harder.
  * debian/control:
- Add iproute2 to Depends.
- Add python3-netifaces to Depends, Build-Depends.

 -- Mathieu Trudel-Lapierre   Mon, 04 Jun 2018
14:45:26 -0400

** Changed in: netplan.io (Ubuntu)
   Status: In Progress => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  Fix Released
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Also affects: netplan.io (Ubuntu)
   Importance: Undecided
   Status: New

** Changed in: netplan.io (Ubuntu)
   Status: New => In Progress

** Changed in: netplan.io (Ubuntu)
 Assignee: (unassigned) => Mathieu Trudel-Lapierre (cyphermox)

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in netplan.io package in Ubuntu:
  In Progress
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

Re: [Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Ryan Harper]

"This is why I added
cloud-init to affected packages -- cloud-init should not be second-
guessing the network layer and attempting to do renames / to run"

There is no second guessing.  In the case where we have no network
config, there is no renaming; we accept whatever name is given.

If the config passed to cloud-init includes a name for an interface, then
cloud-init applies that name; this is the MAAS scenario;

Given the unreliable nature of udev w.r.t naming (see this very bug)
The *only* way for cloud-init to ensure that a directive to name an interface
matches the config (also note cloud-init accepts network config in
various formats
not just netplan) is to handle naming if requested directly, precisely
due to this bug.

If we fix systemd-udevd to allow renames of interfaces reliably that helps
most cases where udevd runs.  For the remaining cases where udev doesn't run,
containers for example, cloud-init will still need to use iproute2 to
set an interface
name if requested.


On Fri, May 25, 2018 at 8:55 AM, Mathieu Trudel-Lapierre
 wrote:
> netplan changes are available in git:
>
> Daniel's patch to write udev rules (SRU material):
> https://github.com/CanonicalLtd/netplan/commit/b0c51bfa8ba8b898a9feaed9cd7d8790d147d35d
>
> Daniel's patch + dropping replug code + rework 'netplan apply' (code for 
> cosmic); in progress for upload to cosmic:
> https://github.com/CanonicalLtd/netplan/tree/live-rename
>
> --
> You received this bug notification because you are subscribed to
> netplan.
> Matching subscriptions: netplan
> https://bugs.launchpad.net/bugs/1770082
>
> Title:
>   systemd-networkd not renaming devices on boot
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: 

Re: [Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Ryan Harper]

On Fri, May 25, 2018 at 9:09 AM, Mathieu Trudel-Lapierre
 wrote:
> I'm not completely sure where the code lives in cloud-init; it looks a
> bit like what's in:
>
> cloudinit/net/netplan.py
>
> But the code does read as though it should not be running 'udevadm test-
> builtin net_setup_link'. However, deploying a system with MAAS shows a
> /etc/cloud/cloud.cfg.d/50-curtin-networking.cfg file, which contains
> data referring to the network interfaces:
>
> network:
>   config:
>   - id: ens6
> mac_address: 52:54:00:31:27:2c
> mtu: 1500
> name: ens6
> subnets:
> - address: 10.3.99.13/24
>   gateway: 10.3.99.1
>   type: static
> type: physical
>   version: 1
>
> With that file in place, any netplan config renaming the interface to
> something other than ens6 will see the interface being renamed *again*
> back to ens6, when that file is removed, this behavior does not appear.

Cloud-init only renders network config once per-instance.  If a user
wanted to modify this to something else (and we fix udev to handle
.link file name changes) then there won't be any additional renames.
Even without the change, cloud-init isn't going to call udevadm test-builtin
more than once per-instance.


>
> --
> You received this bug notification because you are subscribed to
> netplan.
> Matching subscriptions: netplan
> https://bugs.launchpad.net/bugs/1770082
>
> Title:
>   systemd-networkd not renaming devices on boot
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 

Re: [Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Ryan Harper]

On Fri, May 25, 2018 at 8:48 AM, Mathieu Trudel-Lapierre
 wrote:
> My recommended course of action for cosmic:
>
>  - drop udevadm (net_setup_link) call from cloud-init

This will still be needed in the case that we have a network config with names.
Cloud-init runs after udev coldplug, so any .link files would have to already be
present.  We're generating a network config post coldplug, but pre networking.

Now, if after we netplan generate, if links could be applied or if
networkd kicks udev
to rename interfaces, that's fine.  But it doesn't do that today.

>  - drop set-name "renaming" from cloud-init / maas

I've already replied, but this is required for instances where udev
doesn't run but
network config has been provided.

>  - drop replug code in netplan; replace with proper .link code,
possibly call to net_setup_link.

+1

>  - maybe write udev rule for renaming in netplan (think belt and
suspenders)

? Instead of the .link files?

>  - make sure .link files are written correctly by netplan for renaming
>  - patch systemd to not second-guess renaming from .link files

+1

>
> --
> You received this bug notification because you are subscribed to
> netplan.
> Matching subscriptions: netplan
> https://bugs.launchpad.net/bugs/1770082
>
> Title:
>   systemd-networkd not renaming devices on boot
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

My recommended course of action for SRUs:

 - write udev .rules files from netplan to enforce renaming
 - drop udevadm (net_setup_link) call from cloud-init

The above should be sufficient and non-intrusive enough for SRU. The
tasks for cosmic are additional changes to clean up the behavior for the
future.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

I'm not completely sure where the code lives in cloud-init; it looks a
bit like what's in:

cloudinit/net/netplan.py

But the code does read as though it should not be running 'udevadm test-
builtin net_setup_link'. However, deploying a system with MAAS shows a
/etc/cloud/cloud.cfg.d/50-curtin-networking.cfg file, which contains
data referring to the network interfaces:

network:
  config:
  - id: ens6
mac_address: 52:54:00:31:27:2c
mtu: 1500
name: ens6
subnets:
- address: 10.3.99.13/24
  gateway: 10.3.99.1
  type: static
type: physical
  version: 1

With that file in place, any netplan config renaming the interface to
something other than ens6 will see the interface being renamed *again*
back to ens6, when that file is removed, this behavior does not appear.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Patch added: "draft systemd patch to skip should_rename(), adds extra debug 
info"
   
https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/1770082/+attachment/5144272/+files/debug.patch

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

My recommended course of action for cosmic:

 - drop udevadm (net_setup_link) call from cloud-init
 - drop set-name "renaming" from cloud-init / maas
 - drop replug code in netplan; replace with proper .link code, possibly call 
to net_setup_link.
 - maybe write udev rule for renaming in netplan (think belt and suspenders)
 - make sure .link files are written correctly by netplan for renaming
 - patch systemd to not second-guess renaming from .link files

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

netplan changes are available in git:

Daniel's patch to write udev rules (SRU material):
https://github.com/CanonicalLtd/netplan/commit/b0c51bfa8ba8b898a9feaed9cd7d8790d147d35d

Daniel's patch + dropping replug code + rework 'netplan apply' (code for 
cosmic); in progress for upload to cosmic:
https://github.com/CanonicalLtd/netplan/tree/live-rename

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

** Also affects: cloud-init (Ubuntu)
   Importance: Undecided
   Status: New

** Changed in: netplan
   Status: Incomplete => Confirmed

** Changed in: cloud-init (Ubuntu)
   Status: New => Confirmed

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mathieu Trudel-Lapierre]

There are a couple of pieces in play here.

One aspect is that we don't really want to write the .link files for
systemd-networkd to /lib or get anything from /run into the initrd --
that defeats the purpose of netplan's config being dynamic.

The second aspect is that depending on how the systems are deployed
(MAAS? cloud images? openstack? VM? hardware?) changes the behavior a
bit. Some USB-based network interfaces are simply "unaffected" and
rename just fine at boot. Other systems may be renaming the interfaces,
but seeing cloud-init re-change the name *back* to the previous value
(this has been observed for MAAS-deployed systems). This is why I added
cloud-init to affected packages -- cloud-init should not be second-
guessing the network layer and attempting to do renames / to run
udevadm, except *maybe* at first boot/when the instance data is created.
I believe it should just leave it alone completely, and let
netplan/systemd/NM deal with the interfaces themselves (if anything
needs to be poked, it's arguably a bug in the backend).

Another issue is that systemd is also second-guessing configuration. If
we set things to be renamed, we do want them to be renamed. Now, this is
a bit more up to discussion upstream, but I do think systemd should
always rename if configuration tells it to (via .link files or udev
rules -- .link files are "cleaner", easier to write). Users renaming
their interfaces manually via 'ip link dev X name Y' are doing so on
their own, and the change would not be persistent across a reboot --
renaming in configuration is expected to work normally.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Confirmed
Status in cloud-init package in Ubuntu:
  Confirmed
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Ryan Harper]

That's not the final conclusion; the issue is still being discussed
upstream; in particular as to why only the interface name cannot be
modified at runtime by .link files.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Incomplete
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mike Jonkmans]

As discussed on https://github.com/systemd/systemd/issues/9006

The initrd (!) should have NamePolicy=kernel in
/lib/systemd/network/99-default.link

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Incomplete
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

Re: [Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Ryan Harper]

We really only want to be allowed to rename interfaces that have requested it.
This means other interfaces which do not have a set-name directive
will have the kernel name.

On Thu, May 24, 2018 at 6:56 AM, Mike Jonkmans
<1770...@bugs.launchpad.net> wrote:
> Things seem to work when i change the NamePolicy in the initrd file 
> /lib/systemd/network/99-default.link to
> `NamePolicy=kernel`
>
> Leaving out the other NamePolicy-options, prevents renaming in the initrd 
> fase.
> Which allows later renames from .link files/netplan.
>
> --
> You received this bug notification because you are subscribed to
> netplan.
> Matching subscriptions: netplan
> https://bugs.launchpad.net/bugs/1770082
>
> Title:
>   systemd-networkd not renaming devices on boot
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Incomplete
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp

[Touch-packages] [Bug 1770082] Re: systemd-networkd not renaming devices on boot

[Mike Jonkmans]

Things seem to work when i change the NamePolicy in the initrd file 
/lib/systemd/network/99-default.link to
`NamePolicy=kernel`

Leaving out the other NamePolicy-options, prevents renaming in the initrd fase.
Which allows later renames from .link files/netplan.

-- 
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1770082

Title:
  systemd-networkd not renaming devices on boot

Status in netplan:
  Incomplete
Status in systemd package in Ubuntu:
  Confirmed

Bug description:
  === systemd issue ===

  Renaming devices doesn't seem to work.

  If I disable all other network configuration and create
  /etc/systemd/network/10-network.link with:

  [Match]
  MACAddress=52:54:00:c1:c9:bb

  [Link]
  Name=myiface3

  I expect this to cause the device with that MAC address to be renamed
  to  myiface3. However, when I reboot, I instead see:

  $ ip l
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  2: ens3:  mtu 1500 qdisc noop state DOWN mode DEFAULT 
group default qlen 1000
  link/ether 52:54:00:c1:c9:bb brd ff:ff:ff:ff:ff:ff

  The device is not renamed.

  This link file is pretty much identical to Example 2 in
  https://www.freedesktop.org/software/systemd/man/systemd.link.html.

  The renaming does work if I boot with net.ifnames=0, and oddly, it
  also works if I unbind the device and rebind it as netplan apply does.
  No setting of NamePolicy seems to help.

  === Original Bug ==

  'set-name:' doesn't change the name of a network interface on boot, it
  only works when you do netplan apply.

  Say I take this 50-cloud-init.yaml file:

  # This file is generated from information provided by
  # the datasource.  Changes to it will not persist across an instance.
  # To disable cloud-init's network configuration capabilities, write a file
  # /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg with the following:
  # network: {config: disabled}
  network:
  version: 2
  ethernets:
  ens3:
  dhcp4: true
  match:
  macaddress: 52:54:00:de:bd:f6
  set-name: ens3

  Say I change set-name to 'myiface3' and reboot. I expect that the
  device will be called myiface3 and brought up fine with dhcp. However,
  instead I see:

  $ ip a
  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  2: ens3:  mtu 1500 qdisc noop state DOWN group default 
qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff

  The name has not been changed, and the device has not been brought up.

  If I run netplan apply however, I see the following:

  1: lo:  mtu 65536 qdisc noqueue state UNKNOWN group 
default qlen 1000
  link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
  inet 127.0.0.1/8 scope host lo
     valid_lft forever preferred_lft forever
  inet6 ::1/128 scope host
     valid_lft forever preferred_lft forever
  3: myiface3:  mtu 1500 qdisc fq_codel state 
UP group default qlen 1000
  link/ether 52:54:00:de:bd:f6 brd ff:ff:ff:ff:ff:ff
  inet 192.168.122.151/24 brd 192.168.122.255 scope global dynamic myiface3
     valid_lft 3575sec preferred_lft 3575sec
  inet6 fe80::5054:ff:fede:bdf6/64 scope link
     valid_lft forever preferred_lft forever

  So names are successfully changed with netplan apply.

  This seems to be some udev-related timing or priority issue that I'm
  still trying to hunt down.

  This breaks some forms of migration in certain cloud environments.

To manage notifications about this bug go to:
https://bugs.launchpad.net/netplan/+bug/1770082/+subscriptions

-- 
Mailing list: https://launchpad.net/~touch-packages
Post to : touch-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~touch-packages
More help   : https://help.launchpad.net/ListHelp