[nmglug] LINUX Software RAID recovery event?

[Gary Sandine]

On Fri, Jan 11, 2008 at 09:02:33AM -0700, Aaron wrote:
[..]
> Things look OK now AFAIK:
> 
> gromit:/var/log# cat /proc/mdstat
> Personalities : [raid6] [raid5] [raid4]
> md0 : active raid5 sda2[0] sdd2[3] sdc2[2] sdb2[1]
>       585954432 blocks level 5, 64k chunk, algorithm 2 [4/4] [UUUU]
>      
> unused devices: <none>
> 
> Any advice about how to investigate what happened?  Is it possible
> that the RAID failed in some sense, but then re-build itself?  If
> so, is there a way to figure out if the failure was limited to one
> disk...  so I can replace it as a preventative measure?

I have seen software RAID rebuilds occur due to unclean shutdowns
and due to power supply problems (i.e. a disk temporarily loses
power).  

I would look for evidence of disk (or other) problems prior to the
log messages you quoted:

  Jan  6 01:06:02 gromit mdadm: RebuildStarted event detected on md device /dev/md0

Regards,
-- 
Gary Sandine <gars at laclinux.com>