Interesting possible XFS crash condition

[Emmanuel Florac]

Le Wed, 20 Oct 2010 01:13:19 -0500
Shawn Usry <shawn at dolphinlogic.com> écrivait:

> The latter statement and observations lead me to believe that perhaps 
> this was simply a yucky controller that was failing under heavy
> I/O.   

I've set up quite a lot of those RAID cards (about 100) and there is a
significant failure rate on these (much higher than the newer 9650). I
had both cases of bad controller RAM and CPU overheating several times.

Try unmounting the filesystem and start a RAID verify:

tw_cli /cX/uY start verify

This will generate high IO. Check dmesg for controller errors. Try
remounting after a couple of hours of verification. If the controller
is fried, it most probably fail, but shouldn't crash the system.

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |	<eflorac at intellique.com>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------