[OpenAFS] Cache size problems on 1.4.1, Solaris 8,9

[Kevin Hildebrand]

A little more info - apparently I was mistaken - this is only affecting 
Solaris 8 machines, the Solaris 9 machines are fine.

Also, one can duplicate the problem by running something like this 
somewhere in AFS:

find . -exec grep stupid {} \;

Thanks,

Kevin

On Wed, 30 Aug 2006, Kevin Hildebrand wrote:

>
> Hello, lately we've been seeing problems with our cache partitions 
> overfilling on our Solaris 8 and 9 machines running 1.4.1.
>
> For example:
>
> %/usr/vice/etc/fs getcache
> AFS using 448078 of the cache's available 877506 1K byte blocks.
>
> %df -k /usr/vice/cache
> Filesystem            kbytes    used   avail capacity  Mounted on
> /dev/dsk/c0t0d0s4    1018382 1012346       0   100%    /usr/vice/cache
>
> %cat /usr/vice/etc/cacheinfo
> /afs:/usr/vice/cache:877506
>
>
> This problem seems to have started when we switched to 1.4.1.  (I believe we 
> jumped versions to get to 1.4.1, so the problem may have appeared in an 
> earlier version...)
>
> The only solution we've found is to reboot the troubled machine, after which 
> the cache behaves normally.  I have not yet been able to determine a set of 
> conditions that cause the cache to overfill, we've just been kicking the 
> machines as we find them.
>
> Any clues?
>
> Thanks,
>
> Kevin Hildebrand
> University of Maryland, College Park
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
>