On 4/28/2011 2:07 PM, Alexander Farber wrote:
> Hello,
>
> since weeks I was ignoring this warning at my CentOS 5.6/64 bit machine -
>
> /etc/cron.weekly/99-raid-check:
> WARNING: mismatch_cnt is not 0 on /dev/md0
>
> in the hope that the software RAID will slowly repair itself.
>
> I also had executed "echo 100000> /proc/sys/dev/raid/speed_limit_max"
> on the advice from the mailing list.
>
> But now my web server is offline - I had to boot it remotely with rescue system.
>
> Does anybody please have an advice what commands to run
> and do you think it is a RAID problem at all?
>
A 'cat /proc/mdstat' should show the state of the raid mirroring. I
don't see anything that would explain not booting, though. Raid1 works
normally even when only one member is available and should continue to
work while rebuilding. Maybe the problem that caused the mismatch has
corrupted the drive the system normally boots.
--
Les Mikesell
lesmikesell at gmail.com