Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Hi, m...@linux.it (Marco d'Itri) writes: > On May 25, Timo Juhani Lindfors wrote: > >> From this shell I see that the right NIC (i.e. the one with MAC of >> 00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I > Can you still reproduce this with 204-9? I no longer have access to the test setup so I can't test this unfortunately. -Timo -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
On May 25, Timo Juhani Lindfors wrote: > From this shell I see that the right NIC (i.e. the one with MAC of > 00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I Can you still reproduce this with 204-9? -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
m...@linux.it (Marco d'Itri) writes: > Please report the output of "ip link". At what time? When I'm in the beginning of do_nfsmount? When I workaround the problem by using "net0" instead of "eth0" I see oper@watt:~$ ip link 1: lo: mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth2: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:1b:21:a1:d7:83 brd ff:ff:ff:ff:ff:ff 3: net0: mtu 1500 qdisc mq state UP qlen 1000 link/ether 00:25:90:36:c0:d2 brd ff:ff:ff:ff:ff:ff 4: eth3: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:1b:21:a1:d7:82 brd ff:ff:ff:ff:ff:ff 5: eth1: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:25:90:36:c0:d3 brd ff:ff:ff:ff:ff:ff when the system is fully booted up. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Please report the output of "ip link". -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
m...@linux.it (Marco d'Itri) writes:
> Did you rebuild the initramfs as explained in README.Debian?
> Search for NEED_PERSISTENT_NET.
Yes it detected the scenario automatically since fstab has
/dev/nfs / nfserrors=remount-ro 0 0
which is matched by
/usr/share/initramfs-tools/hooks/udev: if egrep -q
'^/dev/(nbd|nd[a-z]|etherd/e)[^[:space:]]+[[:space:]]+/[[:space:]]' /etc/fstab;
then
The right 70-persistent-net.rules is inside the initramfs image:
oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout
lib/udev/rules.d/70-persistent-net.rules
# lindi: ethernet port #1 (BIOS boots using this)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:25:90:36:c0:d2", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth0"
# PCI device 0x8086:0x10ec (ixgbe)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:1b:21:a1:d7:83", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth2"
# PCI device 0x8086:0x10ec (ixgbe)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:1b:21:a1:d7:82", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth3"
# PCI device 0x8086:0x10c9 (igb)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:25:90:36:c0:d3", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth1"
53853 blocks
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
On May 25, Timo Juhani Lindfors wrote: > oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout > lib/udev/rules.d/70-persistent-net.rules Did you rebuild the initramfs as explained in README.Debian? Search for NEED_PERSISTENT_NET. -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Package: initramfs-tools
Version: 0.98.8
Severity: normal
Description
===
On a network booting machine with
oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout
lib/udev/rules.d/70-persistent-net.rules
# lindi: ethernet port #1 (BIOS boots using this)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:25:90:36:c0:d2", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth0"
# PCI device 0x8086:0x10ec (ixgbe)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:1b:21:a1:d7:83", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth2"
# PCI device 0x8086:0x10ec (ixgbe)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:1b:21:a1:d7:82", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth3"
# PCI device 0x8086:0x10c9 (igb)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*",
ATTR{address}=="00:25:90:36:c0:d3", ATTR{dev_id}=="0x0", ATTR{type}=="1",
KERNEL=="eth*", NAME="eth1"
53853 blocks
and
$ grep DEVICE= /etc/initramfs-tools/initramfs.conf
DEVICE=eth0
I see that ipconfig gets stuck at bootup trying to send DHCP requests
over a NIC whose MAC is not 00:25:90:36:c0:d2.
Troubleshooting
===
I placed a call to "/bin/sh" just before configure_networking call in
oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout
scripts/nfs | head -n15
# NFS filesystem mounting -*- shell-script -*-
# FIXME This needs error checking
retry_nr=0
# parse nfs bootargs and mount nfs
do_nfsmount()
{
configure_networking
# get nfs root from dhcp
if [ "x${NFSROOT}" = "xauto" ]; then
# check if server ip is part of dhcp root-path
>From this shell I see that the right NIC (i.e. the one with MAC of
00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I
can see in source code of udev that it sometimes generates such names
but I don't fully understand the logic behind that. I did
rmmod ixgbe
rmmod igb
modprobe igb
modprobe ixgbe
and suddenly 00:25:90:36:c0:d2 was correctly named eth0.
The machine has 36 SATA disks and it takes a while to detect
them. Maybe there's a race condition somewhere?
As a workaround I'm now calling my 00:25:90:36:c0:d2 NIC "net0"
instead of "eth0".
-- Package-specific info:
-- initramfs sizes
-rw-r--r-- 1 root root 9.1M May 23 16:01 /boot/initrd.img-2.6.32-5-amd64
-- /proc/cmdline
root=/dev/nfs initrd=watt/initrd.img nfsroot=10.8.8.8:/tftpboot/watt ip=dhcp
sysrq_always_enabled break=top BOOT_IMAGE=watt/vmlinuz
-- resume
RESUME=UUID=d2e19371-27b1-4826-9d93-dd0f4e1f0ea3
-- /proc/filesystems
-- lsmod
Module Size Used by
loop 11799 0
snd_pcm60503 0
snd_timer 15582 1 snd_pcm
snd46446 2 snd_pcm,snd_timer
soundcore 4598 1 snd
snd_page_alloc 6249 1 snd_pcm
psmouse49937 0
joydev 8459 0
pcspkr 1699 0
serio_raw 3752 0
i2c_i8017830 0
evdev 7352 4
ioatdma34876 32
i2c_core 15819 1 i2c_i801
button 4650 0
processor 29935 16
nfs 241114 1
lockd 57603 1 nfs
fscache29834 1 nfs
nfs_acl 2031 1 nfs
auth_rpcgss33476 1 nfs
sunrpc161541 9 nfs,lockd,nfs_acl,auth_rpcgss
sd_mod 29921 0
crc_t10dif 1276 1 sd_mod
usbhid 33292 0
hid63257 1 usbhid
uhci_hcd 18521 0
ehci_hcd 32081 0
ixgbe 106622 0
usbcore 122386 4 usbhid,uhci_hcd,ehci_hcd
mpt2sas87773 0
scsi_transport_sas 19865 1 mpt2sas
mdio3096 1 ixgbe
scsi_mod 122149 3 sd_mod,mpt2sas,scsi_transport_sas
nls_base6377 1 usbcore
igb77959 0
dca 3761 3 ioatdma,ixgbe,igb
thermal11674 0
thermal_sys11942 2 processor,thermal
-- /etc/initramfs-tools/modules
-- /etc/kernel-img.conf
# Kernel image management overrides
# See kernel-img.conf(5) for details
do_symlinks = yes
do_bootloader = no
do_initrd = yes
link_in_boot = no
-- /etc/initramfs-tools/initramfs.conf
MODULES=most
BUSYBOX=y
KEYMAP=n
COMPRESS=gzip
BOOT=nfs
DEVICE=eth0
NFSROOT=auto
-- /etc/initramfs-tools/update-initramfs.conf
update_initramfs=yes
backup_initramfs=no
-- mkinitramfs hooks
/etc/initramfs-tools/hooks/:
/usr/share/initramfs-tools/hooks:
busybox
dmsetup
keymap
klibc
mdadm
thermal
udev
-- System Information:
Debian Release: 6.0.1
APT prefers stable-updates
APT policy: (500, 'stable-updates'), (500, 'stable')
Architecture: amd64 (x86_64)
Kernel: Linux 2.6.32-5-amd64 (SMP w/16 CPU cores)
Locale: LANG=C, LC_CTYPE=en_GB.UTF-8
