[ovirt-users] qcow2 images corruption

[Yaniv Kaul]

On Feb 7, 2018 7:08 PM, "Nicolas Ecarnot" <nicolas at ecarnot.net> wrote:

Hello,

TL; DR : qcow2 images keep getting corrupted. Any workaround?

Long version:
This discussion has already been launched by me on the oVirt and on
qemu-block mailing list, under similar circumstances but I learned further
things since months and here are some informations :

- We are using 2 oVirt 3.6.7.5-1.el7.centos datacenters, using CentOS
7.{2,3} hosts
- Hosts :
  - CentOS 7.2 1511 :
    - Kernel = 3.10.0 327
    - KVM : 2.3.0-31
    - libvirt : 1.2.17
    - vdsm : 4.17.32-1
  - CentOS 7.3 1611 :
    - Kernel 3.10.0 514
    - KVM : 2.3.0-31
    - libvirt 2.0.0-10
    - vdsm : 4.17.32-1


All are somewhat old releases. I suggest upgrading to the latest RHEL and
qemu-kvm bits.

Later on, upgrade oVirt.
Y.

- Our storage is 2 Equallogic SANs connected via iSCSI on a dedicated
network
- Depends on weeks, but all in all, there are around 32 hosts, 8 storage
domains and for various reasons, very few VMs (less than 200).
- One peculiar point is that most of our VMs are provided an additional
dedicated network interface that is iSCSI-connected to some volumes of our
SAN - these volumes not being part of the oVirt setup. That could lead to a
lot of additional iSCSI traffic.

>From times to times, a random VM appears paused by oVirt.
Digging into the oVirt engine logs, then into the host vdsm logs, it
appears that the host considers the qcow2 image as corrupted.
Along what I consider as a conservative behavior, vdsm stops any
interaction with this image and marks it as paused.
Any try to unpause it leads to the same conservative pause.

After having found (https://access.redhat.com/solutions/1173623) the right
logical volume hosting the qcow2 image, I can run qemu-img check on it.
- On 80% of my VMs, I find no errors.
- On 15% of them, I find Leaked cluster errors that I can correct using
"qemu-img check -r all"
- On 5% of them, I find Leaked clusters errors and further fatal errors,
which can not be corrected with qemu-img.
In rare cases, qemu-img can correct them, but destroys large parts of the
image (becomes unusable), and on other cases it can not correct them at all.

Months ago, I already sent a similar message but the error message was
about No space left on device (https://www.mail-archive.com/
qemu-block at gnu.org/msg00110.html).

This time, I don't have this message about space, but only corruption.

I kept reading and found a similar discussion in the Proxmox group :
https://lists.ovirt.org/pipermail/users/2018-February/086750.html

https://forum.proxmox.com/threads/qcow2-corruption-after-
snapshot-or-heavy-disk-i-o.32865/page-2

What I read similar to my case is :
- usage of qcow2
- heavy disk I/O
- using the virtio-blk driver

In the proxmox thread, they tend to say that using virtio-scsi is the
solution. Having asked this question to oVirt experts (
https://lists.ovirt.org/pipermail/users/2018-February/086753.html) but it's
not clear the driver is to blame.

I agree with the answer Yaniv Kaul gave to me, saying I have to properly
report the issue, so I'm longing to know which peculiar information I can
give you now.

As you can imagine, all this setup is in production, and for most of the
VMs, I can not "play" with them. Moreover, we launched a campaign of
nightly stopping every VM, qemu-img check them one by one, then boot.
So it might take some time before I find another corrupted image.
(which I'll preciously store for debug)

Other informations : We very rarely do snapshots, but I'm close to imagine
that automated migrations of VMs could trigger similar behaviors on qcow2
images.

Last point about the versions we use : yes that's old, yes we're planning
to upgrade, but we don't know when.

Regards,

-- 
Nicolas ECARNOT
_______________________________________________
Users mailing list
Users at ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20180208/dc498094/attachment.html>