Am Mittwoch, 9. März 2005 01:26 schrieb Ed Murray:
> Hi I am using XFS on linux kernel 2.6.10 with a 3ware raid 0 7006
> controller. Three days ago my server crashed. After rebooting I found
> log entries such as the following in my syslog. I booted up a system
> rescue cd and ran xfs_repair. It seemed to work. I then ran xfs_check
> and it ran without errors. I have upgraded the kernel to 2.6.11 and I
> have disabled 4K stacks. However I seem to still be experiencing
> corruption.
>
> My question is: Can I stop this from reoccurring or is my file system
> hosed and I should reformat and start again?
No. You should run xfs_repair /dev/your_hdd.
You have to dismount the filesystem to do that.
To stop further corruption you will have to make sure, that your
controller syncs *any* write-cached data to harddisk.
I think your controller does argressive write caching. If you use an
controller with hardware caches you have to flusch the cache before
rebooting. Also a UPS ist recommended for such controllers...
Jojo
--
Jürgen Sauer - AutomatiX GmbH, +49-4209-4699, jojo@xxxxxxxxxxxx **
** Das Linux Systemhaus - Service - Support - Server - Lösungen **
** http://www.automatix.de ICQ: #344389676 **
OpenOffice erhalten Sie hier kostenfrei http://de.openoffice.org/
pgpE9DJxXV7ya.pgp
Description: PGP signature
|