Re: XFS filesystem on EC2 instance corrupts and shuts down

[Shrinath M <shrinath.m@xxxxxxxxxx>]

Thanks Ben, Dave and Eric.Â

Eric,Â

>>but I am wondering if there might be more information before this which is not in your trimmed logs.

No, this was the first entry every time we have it in /var/log/messages. dmesg also holds the same. After reboot, it simply fixes without anyone doing anything.

The Linux we are running is definitely amazon baked one, looks like this -Â

$~: uname -a Linux ip-100-0-100-1 3.2.34-55.46.amzn1.x86_64 #1 SMP Tue Nov 20 10:06:15 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux

Â- dmesg shows something like this after repairing/rebooting -Â

[ Â Â8.414176] SGI XFS with ACLs, security attributes, realtime, large block/inode numbers, no debug enabled

[ Â Â8.415342] SGI XFS Quota Management subsystem

[ Â Â8.417664] XFS (md0): Mounting Filesystem

[ Â Â8.771553] XFS (md0): Starting recovery (logdev: internal)

[ Â Â9.977325] XFS (md0): Ending recovery (logdev: internal)

Check the first line there, it says no debug enabled. How good/bad is this debug mode in production environments? We are not getting any corruption in our local/test environments, in production, we are getting it once on every third day.

Dave,Â

You say unlinked inode list, but if that, it should have an entry in /var/log/messages, right?

Anyway, how can we create this situation? By forcing multiple processes to write/delete files from small disk? Since we are still unaware of what is causing this issue, reproducing it in local/production environment is just shooting in dark... :(

Does turning up the error level affect the data in any way? Or is it *just* detailed good logging while being sensitive to all small errors?

Really appreciate the support that you devs are giving which really is the job of AWS support... I so wish they had some helpful and knowledgeable people in support.

On Thu, Mar 14, 2013 at 5:12 AM, Dave Chinner <david@xxxxxxxxxxxxx> wrote:

On Wed, Mar 13, 2013 at 01:56:35PM -0500, Eric Sandeen wrote:
> XFS (md0): xfs_iunlink_remove: xfs_itobp() returned error 117.

Corrupted unlinked inode list. You need to run xfs_repair to fix
this.

Chers,

Dave.
--
Dave Chinner
david@xxxxxxxxxxxxx

--
Regards

Shrinath.M