Segmentation fault

[XZed]

Hello,

I found 2/3 topics, on this forum, about this error and some URL (http://imil.net/wp/?p=230), but don't manage to resolve it.

Many ppl quote a faulty rtl8139 driver.

As adviced on this forum, i tried e1000 : the same error.

And tried too ne2k_pci : the same error.

When does it happen :

My tests showed that it happens during a reboot initialized directly on the VM or by the proxmox (for example : backup task with stop mode).

I didn't initially took a screenshot, but as i noticed it's recurrent i took this screenshot and think the error is similar each time :

http://fr.tinypic.com/view.php?pic=rrnxwp&s=4

By the way, it's starting giving me strange errors on backup logs :

209: Oct 01 01:07:11 INFO: final sync finished (87 seconds)
209: Oct 01 01:07:11 INFO: restarting vps
209: Oct 01 01:07:12 INFO: open /dev/kvm: No such file or directory
209: Oct 01 01:07:12 INFO: Could not initialize KVM, will disable KVM support
209: Oct 01 01:07:12 INFO: vps is online again after 158 seconds

Or

209: Oct 02 01:05:30 INFO: first sync finished (327 seconds)
209: Oct 02 01:05:30 INFO: stopping vps
209: Oct 02 01:15:30 INFO: wait failed - got timeout
209: Oct 02 01:15:35 ERROR: Backup of VM 209 failed - command 'qm shutdown 209 && qm wait 209 600' failed with exit code 255

I also had one "kernel panic" / "vfs syncing" error...

For the moment, i setup the backup with snapshots mode and will see if the problem is specific to reboot command (what i think...).

I found some results about ACPI/APM settings but nothing accurate..

Here are my settings :

Proxmox v1.3
Dell PE T100
Only 1 guest KVM with 1 CPU "allocated".
NIC drivers tested : e1000, rtl8139, ne2k_pci


If someone could help me..thank you very much..

Sincerely,

 

[dietmar]

209: Oct 01 01:07:12 INFO: open /dev/kvm: No such file or directory
209: Oct 01 01:07:12 INFO: Could not initialize KVM, will disable KVM support

Seems kvm module is not loaded at all - why?

 

[XZed]

Seems kvm module is not loaded at all - why?

What i didn't understand : this message appeared only once !

I don't know in which way to search for all these problems..

Thank you

 

[dietmar]

I also had one "kernel panic" / "vfs syncing" error...

Is that on the host or inside the VM? And where do you get that 'Segmentation fault' - exactly?

 

[XZed]

Is that on the host or inside the VM? And where do you get that 'Segmentation fault' - exactly?

Hello,

let me accurate these informations :

the two errors you quote appeared inside the VM.

The log about disabled kvm only (???) appeared once on backup logs.

By the way, i had the successfull backup results this morning, using snapshot mode :

vzdump --quiet --snapshot --compress --dumpdir /backup --mailto xxxx@yyyy.zz 209
209: Oct 03 01:00:02 INFO: Starting Backup of VM 209 (qemu)
209: Oct 03 01:00:02 INFO: status = running
209: Oct 03 01:00:02 INFO: creating lvm snapshot of /dev/mapper/pve-data ('/dev/pve/vzsnap')
209: Oct 03 01:00:04 INFO: Logical volume "vzsnap" created
209: Oct 03 01:00:04 INFO: mounting lvm snapshot
209: Oct 03 01:00:04 INFO: creating archive '/backup/vzdump-209.dat' (/mnt/vzsnap/images/209)
209: Oct 03 01:00:04 INFO: qemu-server.conf
209: Oct 03 01:00:04 INFO: vm-102-disk.qcow2
209: Oct 03 01:02:30 INFO: Total bytes written: 3626659840 (3.4GiB, 24MiB/s)
209: Oct 03 01:02:30 INFO: file size 895MB
209: Oct 03 01:02:32 INFO: Logical volume "vzsnap" successfully removed
209: Oct 03 01:02:32 INFO: Finished Backup of VM 209 (00:02:30)

I didn't yet tried restoring such a snapshot, but i'm trusting previous proxmox staff member's advices about the reliability of snapshot mode !

But, i never liked to "bypass" problems and really need to setup a "stop" backup without fearing a freeze of my VM.

Once again, thank you very much for you precious help !

Sincerely,