Arch Linux

Stack

Member

Registered: 2012-04-22

Posts: 13

Timeout issue and LVM volume not found while early boot.

I got LVM volume not found issue while early boot and when the system is "properly set up". I.e. the system was able to boot successfully.

Can't catch whether or not the issue is considered to be solved. There were news: big changes to LVM2 in testing - https://mailman.archlinux.org/pipermail … 31906.html So, the system was revised, similar issue was found before revision ( https://bugs.archlinux.org/task/30966#comment100158 ).

There are similar topics:
https://bbs.archlinux.org/viewtopic.php?id=145714
https://bbs.archlinux.org/viewtopic.php?id=148385

The main question - how this issue may be solved before reboot, when a system is still on the run?

I solved it. Unfortunately, now I do not have a whole free day and hardware resources (I have no time) to try to reproduce the issue and so on...

Initially I had clean Arch install, CD 2012-12... was used. Root FS is located on

HDD x2
|
+- DM RAID0
    |
    +- LVM with root FS
        |
        +- and exist active spare volume for root FS 

Full system upgrade and reboot was done.
Everything was still OK.

Several Gb of software installation was done (~1,5Gb of packages and ~5-7Gb of staff to be extracted while installation, this was mirrored by LVM at spare volume).

Subsequent reboot - that's it - I had timeout + device not found issue:

:: running early hook [udev]
:: running early hook [lvm2]
:: running hook [udev]
:: Triggering uevents...
Waiting 10 seconds for device /dev/mapper/grp_main-vol_rootfs ...
ERROR: device '/dev/mapper/grp_main-vol_rootfs' not found. Skipping fsck.
ERROR: Unable to find root device '/dev/mapper/grp_main-vol_rootfs'.
You are being dropped to a recovery shell
    Type 'exit' to try and continue booting
sh: can't access tty; job control turned off
[rootfs /]# cat /proc/mdstat
Personalities : [raid0] [raid1] [raid6] [raid5] [raid4]
md3 : active raid1 sda3[0] sdb3[1]
      12345678 blocks super 1.2 [2/2] [UU]

md2 : active raid0 sda2[0] sdb2[1]
      12345678 blocks super 1.2 512k chunks

unused devices: <none>
[rootfs /]#

HDD at the moment was doing continuously 'tsrt-tsrt-tsrt... ... tsrt-tsrt-tsrt... ... tsrt-tsrt-tsrt... ...' for a long time. After a while it stop doing 'tsrt', the noises become usual run idle sounds. Only one HDD. It was checked by data cable pull and push.  The trick is dangerous for hardware, but possible in my hardware configuration. Power supply was strong enough, the HDD had SATA and Molex power outlets, both were used. It was not power related issue.

After a cup of tee (long waiting), without any cable touching, when the system was allowed to boot without intervention, when HDD become quiet, 'exit' command may be done from the recovery shell and the system will successfully resume boot. But the issue will not go away, reboot do not affect it.

After spare volume was dropped and new spare volume was created the issue disappeared.

PartedMagic live CD do not have the issue. It's on-screen log indicating, it is simply waiting for the device to become active.

How this issue may be solved before reboot, when a system is still on the run?

Last edited by Stack (2013-02-25 15:26:59)