>Ijforis@xxxxxxxxx writes:
>may have found a way to reproduce a file corruption bug and I would
l>ike to know
if >I am seeing something unique to our environment, or if this is a
>problem for everyone.
>
>Summary: when writing to a XFS formated software raid0 partition which
>is > 70% full,
>unmounting, then remounting the partition will show random 4K block file
>corruption in
>files larger than the raid chunk size. We (myself and a coworker) have
>tested 2.6.8-rc2-bk5
>and 2.6.11; both show the same behavior.
Any advice from the XFS team on this?
We have numerous production machines configured very similarly -- 2 Hardware
RAID-5 arrays striped with Software RAID-0 formatted with XFS. I would bet
that, since our machines had their RAIDS > 50 % full, none have been rebooted
(they've been up only several months at the various locations, but the RAIDS
fill up very quickly).
In our case, the main difference is that we're using the 2.6.6 smp kernel
and internal journals. Is this something that might have developed since the
2.6.6 kernel?
Is there any chance this could be connected to any kind of caching by the
drives (or in our case, by the Hardware RAID card)?
Most urgently, is there something we should or CAN do to check our data on a
preventive basis?
Thanks,
Andy Liebman
|