Skip to contentSkip to conversation

xfs_log_force: error 5 returend

DISCUSSION Steffen Knauf 4 participants 4 messages Sep 14, 2009

lore
Share & tools
lore
#0 lore raw

hello,

sometimes, our server crashed with xfs_log_force: error 5.
It isn't reproducible and a xfs_check / xfs_repair don't find some errors.

Perhaps it's a bug?

Version: Linux version 2.6.26-bpo.1-amd64 (Debian 2.6.26-9~bpo40+1) ([email protected]) (gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21

greets

Steffen

Signature · 7 lines
Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3 -
sicherer, schneller und einfacher! http://portal.gmx.net/de/go/chbrowser

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#1 lore raw
Quoted 5 lines · Steffen Knauf

hello,

sometimes, our server crashed with xfs_log_force: error 5.
It isn't reproducible and a xfs_check / xfs_repair don't find some errors.

Quoted 7 lines

Perhaps it's a bug?

Version: Linux version 2.6.26-bpo.1-amd64 (Debian 2.6.26-9~bpo40+1) ([email protected]) (gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21

greets

Steffen

Hello!

Look at http://oss.sgi.com/bugzilla/show_bug.cgi?id=840
Please provide your dmesg output regarding kernel oops message (maybe there is similar issue in your case as I had)
It seems that in my case it was a problem of HBA configuration (using LVM2 and caches) as Christoph Hellwig replied. After proper configuration I had similar issues still, but last Friday I upgraded HBA firmware because I've found that there was an issue with cache mode report (software reported that write caches are disabled but they were still enabled). I hope this will help in my case - for now I had no crash. We'll see within next week (it occurs about once a week).

In your case without oops message it's really hard to tell what's wrong.

Best regards,
Tomasz Kruszona

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#2 lore raw
In reply to Steffen Knauf
Quoted 6 lines · Steffen Knauf

hello,

sometimes, our server crashed with xfs_log_force: error 5.
It isn't reproducible and a xfs_check / xfs_repair don't find some errors.

Perhaps it's a bug?

Error 5 is EIO, and xfs_log_force basically passed the error through from the underlying device. Most likely it is an issue with the underlying storage device.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#3 lore raw
Quoted 11 lines · Christoph Hellwig
Quoted 6 lines · Steffen Knauf

hello,

sometimes, our server crashed with xfs_log_force: error 5.
It isn't reproducible and a xfs_check / xfs_repair don't find some errors.

Perhaps it's a bug?

Error 5 is EIO, and xfs_log_force basically passed the error through from the underlying device. Most likely it is an issue with the underlying storage device.

Also, just to be semantically correct, it's unlikely that this was a "server crash" - most likely you got an IO error, xfs noticed it, and shut itself down for safety.

-Eric

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs