[Logwatch-Devel] RAID errors

Kirk Bauer kirk at kaybee.org
Thu Aug 5 06:32:10 MST 2004


On Thu, 5 Aug 2004, [ISO-8859-2] Pawe³ Go³aszewski wrote:

> I've got in my logs this report:
>
> Aug  4 00:23:49 raid1: Disk failure on sda5, disabling device.
> Aug  4 00:23:49 md: (skipping faulty sda5 )
> Aug  4 00:23:49 md3: no spare disk to reconstruct array! -- continuing in degraded mode
>
> Problem is that it was outside any filter. It's not matched anywhere else
> and is extremely important.
>
> What can we do about that?

That's strange that it isn't reported as a 'kernel' message.  The 'raid'
filter seems to assume that.  Unfortunately, with messages related to
'raid*' and 'md*' those are a lot of "services" to match for.

I actually had thought about writing a filter to parse /proc/mdstat and
report any errors.  But I have limited data to work with... here is what
mine says:

# cat /proc/mdstat
Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdd1[2] sde1[1] sdc1[0]
      97691008 blocks level 5, 32k chunk, algorithm 2 [3/3] [UUU]

unused devices: <none>


-- 
Kirk Bauer <kirk at kaybee.org>
http://linux.kaybee.org | www.autorpm.org | www.logwatch.org



More information about the Logwatch-Devel mailing list