[Gluster-users] Gluster 3.8.13 data corruption

[Mahdi Adnan]

I only got this issue with this one template (CentOS7), i checked other templates and it worked fine.
I didn't quite understand what you need, i have one UUID and three files, one for the img and .meta and the other one is .lease.
I got the output of all three files which represent the VM i created from the template, is this correct ?

https://paste.fedoraproject.org/paste/qzHmK8t-eJHM3hcZBVs5Yw


--

Respectfully
Mahdi A. Mahdi

________________________________
From: Krutika Dhananjay <kdhananj at redhat.com>
Sent: Monday, October 9, 2017 1:59 PM
To: Mahdi Adnan
Cc: Lindsay Mathieson; gluster-users at gluster.org
Subject: Re: [Gluster-users] Gluster 3.8.13 data corruption

OK.

Is this problem unique to templates for a particular guest OS type? Or is this something you see for all guest OS?

Also, can you get the output of `getfattr -d -m . -e hex <path>` for the following two "paths" from all of the bricks:
path to the file representing the vm created off this template wrt the brick. It will usually be $BRICKPATH/xxxx....xx/images/$UUID where $UUID represents the uuid of the vm created from the template.  If im not wrong, there would be two sets of a certain UUID2, UUID2.lease, UUID2.meta. Please  get me the output of the command above for both the uuid files. The one that has more than 2 hard links is the template. When you attach this output, please make that distinction clear, so it will be easier to debug.


-Krutika

On Fri, Oct 6, 2017 at 6:38 PM, Mahdi Adnan <mahdi.adnan at outlook.com<mailto:mahdi.adnan at outlook.com>> wrote:

Hi,

Thank you for your reply.
Lindsay,
Uunfortunately i do not have backup for this template.

Krutika,
The stat-prefetch is already disabled on the volume.

--

Respectfully
Mahdi A. Mahdi

________________________________
From: Krutika Dhananjay <kdhananj at redhat.com<mailto:kdhananj at redhat.com>>
Sent: Friday, October 6, 2017 7:39 AM
To: Lindsay Mathieson
Cc: Mahdi Adnan; gluster-users at gluster.org<mailto:gluster-users at gluster.org>
Subject: Re: [Gluster-users] Gluster 3.8.13 data corruption

Could you disable stat-prefetch on the volume and create another vm off that template and see if it works?

-Krutika

On Fri, Oct 6, 2017 at 8:28 AM, Lindsay Mathieson <lindsay.mathieson at gmail.com<mailto:lindsay.mathieson at gmail.com>> wrote:
Any chance of a backup you could do bit compare with?

Sent from my Windows 10 phone

From: Mahdi Adnan<mailto:mahdi.adnan at outlook.com>
Sent: Friday, 6 October 2017 12:26 PM
To: gluster-users at gluster.org<mailto:gluster-users at gluster.org>
Subject: [Gluster-users] Gluster 3.8.13 data corruption

Hi,

We're running Gluster 3.8.13 replica 2 (SSDs), it's used as storage domain for oVirt.
Today, we found an issue with one of the VMs template, after deploying a VM from this template it will not boot, it stuck at mount the root partition.
We've been using this templates for months now and we did not had any issues with it.
Both oVirt and Gluster logs is not showing any errors or warnings.
I exported the template and tried running on a standalone machine and it did not work.
Do you think this might be a silent data corruption ? Bitrot is not enabled.
Volume info can be found at:
https://paste.fedoraproject.org/paste/f~1jNIObDa2zoG7ATdCdZg

--

Respectfully
Mahdi A. Mahdi


_______________________________________________
Gluster-users mailing list
Gluster-users at gluster.org<mailto:Gluster-users at gluster.org>
http://lists.gluster.org/mailman/listinfo/gluster-users


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20171009/ed9c405a/attachment.html>