Public bug reported:
While running a dist-ugprade on Jammy machines part of a Ceph cluster,
dbus got updated
```
2024-09-25 04:35:25 status triggers-pending dbus:amd64 1.12.20-2ubuntu4.1
2024-09-25 04:35:36 status triggers-pending dbus:amd64 1.12.20-2ubuntu4.1
2024-09-25 04:37:28 trigproc dbus:amd64 1.12.20-2ubuntu4.1 <none>
2024-09-25 04:37:28 status half-configured dbus:amd64 1.12.20-2ubuntu4.1
2024-09-25 04:37:28 status installed dbus:amd64 1.12.20-2ubuntu4.1
```
This I believe trigger a systemd-networkd restart
```
Sep 25 04:36:35 ********* systemd[1]: Started Network Configuration.
```
Which causes interfaces to go down and up.
```
Sep 25 04:36:35 ********* systemd-networkd[605770]: bond0: netdev exists,
using existing without changing its parameters
Sep 25 04:36:35 ********* systemd-networkd[605770]: bond1: netdev exists,
using existing without changing its parameters
Sep 25 04:36:35 ********* systemd-networkd[605770]: br-bond0-3486: netdev
exists, using existing without changing its parameters
Sep 25 04:36:35 ********* systemd-networkd[605770]: br-bond1: netdev exists,
using existing without changing its parameters
Sep 25 04:36:35 ********* systemd-networkd[605770]: br-enp68s0f0: netdev
exists, using existing without changing its parameters
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp129s0f1np1: Link DOWN
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp129s0f1np1: Lost
carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp129s0f0np0: Link DOWN
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp129s0f0np0: Lost
carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: bond1: Lost carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp161s0f1np1: Link DOWN
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp161s0f1np1: Lost
carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp161s0f0np0: Link DOWN
Sep 25 04:36:37 ********* systemd-networkd[605770]: enp161s0f0np0: Lost
carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: bond0: Lost carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: bond0.3484: Lost carrier
Sep 25 04:36:37 ********* systemd-networkd[605770]: bond0.3486: Lost carrier
Sep 25 04:36:41 ********* systemd-networkd[605770]: enp129s0f1np1: Link UP
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp129s0f1np1: Gained
carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp129s0f0np0: Link UP
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp129s0f0np0: Gained
carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp161s0f1np1: Link UP
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp161s0f0np0: Link UP
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp161s0f1np1: Gained
carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: enp161s0f0np0: Gained
carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: bond0: Gained carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: bond0.3484: Gained carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: bond0.3486: Gained carrier
Sep 25 04:36:42 ********* systemd-networkd[605770]: bond1: Gained carrier
```
```
Sep 25 04:36:35 ********* kernel: [18085928.795785] bond1: (slave
enp129s0f1np1): link status definitely down, disabling slave
Sep 25 04:36:36 ********* kernel: [18085929.535826] bond1: (slave
enp129s0f0np0): link status definitely down, disabling slave
Sep 25 04:36:37 ********* kernel: [18085929.919702] bond0: (slave
enp161s0f1np1): link status definitely down, disabling slave
Sep 25 04:36:37 ********* kernel: [18085930.339709] bond0: (slave
enp161s0f0np0): link status definitely down, disabling slave
Sep 25 04:36:41 ********* kernel: [18085934.812301] bond0: (slave
enp161s0f1np1): link status definitely up, 25000 Mbps full duplex
Sep 25 04:36:41 ********* kernel: [18085934.812968] bond0: (slave
enp161s0f0np0): link status definitely up, 25000 Mbps full duplex
Sep 25 04:36:42 ********* kernel: [18085934.828048] bond1: (slave
enp129s0f1np1): link status definitely up, 100000 Mbps full duplex
Sep 25 04:36:42 ********* kernel: [18085934.828491] bond1: (slave
enp129s0f0np0): link status definitely up, 100000 Mbps full duplex
```
This caused a big outage on the Ceph cluster where many placement groups
went unavailable.
It will be good if the dbus upgrade can avoid to restart the dbus
service, to avoid the consequent impact to networking, and leave that to
the operator to do it at its best convenience.
** Affects: dbus (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2083489
Title:
Upgrading dbus tear down network
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/dbus/+bug/2083489/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
