Windows-VM killed at beginning of Backup (VSS/traffic) running on pvetest

[udo]

Hi,
due to problems with an ESX-host I migrate two windows-VMs (2003) to an pve-host, which is running pvetest.

One VM run without trouble - also backup with bacula.
The second VM was killed at the beginning of the backup.
In messages ist only

Feb 14 21:00:49 proxtest4 kernel: vmbr0: port 3(tap120i0) entering disabled state
Feb 14 21:00:49 proxtest4 kernel: vmbr0: port 3(tap120i0) entering disabled state

and due starting VM again:

Feb 14 21:01:04 proxtest4 kernel: device tap120i0 entered promiscuous mode
Feb 14 21:01:04 proxtest4 kernel: vmbr0: port 3(tap120i0) entering forwarding state

In syslog I see also the console closing.
Differences between (backup) running and failed nodes:
the running nodes has only one virtio-disk - the faild node has an ide disk (C because virtio wasn't installable and 4 scsi-disks

running node:

boot: cdn
bootdisk: virtio0
cores: 2
virtio0: ssd:vm-110-disk-1
ide2: local:iso/virtio-win-0.1-52.iso,media=cdrom,size=57126K
memory: 4096
name: termserver
net0: virtio=6A:3B:39:D1:3A:48,bridge=vmbr0
ostype: wxp
sockets: 2

failed node:

bootdisk: ide0
cores: 1
ide0: ssd:vm-120-disk-1,size=12G
scsi0: ssd:vm-120-disk-2,size=25G
scsi1: ssd:vm-120-disk-3,size=25G
scsi2: ssd:vm-120-disk-4,size=40G
scsi3: ssd:vm-120-disk-5,size=6G
ide2: local:iso/virtio-win-0.1-52.iso,media=cdrom,size=57126K
memory: 4096
name: dbhost
net0: virtio=7E:D7:0F:9E:AA:F8,bridge=vmbr0
ostype: wxp
sockets: 2

I will try to create an test-vm also with ide/scsi to see if the issue is reproducable.

Any hint's?

Version:

pve-manager: 2.3-7 (pve-manager/2.3/1fe64d18)
running kernel: 2.6.32-18-pve
proxmox-ve-2.6.32: 2.3-88
pve-kernel-2.6.32-18-pve: 2.6.32-88
pve-kernel-2.6.32-17-pve: 2.6.32-83
lvm2: 2.02.95-1pve2
clvm: 2.02.95-1pve2
corosync-pve: 1.4.4-4
openais-pve: 1.1.4-2
libqb: 0.10.1-2
redhat-cluster-pve: 3.1.93-2
resource-agents-pve: 3.9.2-3
fence-agents-pve: 3.1.9-1
pve-cluster: 1.0-36
qemu-server: 2.3-8
pve-firmware: 1.0-21
libpve-common-perl: 1.0-44
libpve-access-control: 1.0-25
libpve-storage-perl: 2.3-2
vncterm: 1.0-3
vzctl: 4.0-1pve2
vzprocps: 2.0.11-2
vzquota: 3.1-1
pve-qemu-kvm: 1.3-18
ksm-control-daemon: 1.1-1

Udo

 

[udo]

Hi,
if I create an new win2003 VM with ide as C and an scsi-disk, the VM breaks during initialising of the scsi-volume (formating, no fast format).

For me it's looks scsi-related. I will try the driver from lsi and not the build in from windows...

Udo

 

[dietmar]

Ore you can try to use another scsi controller type.

 

[udo]

Ore you can try to use another scsi controller type.

Hi Dietmar,
the same behavior with Megaraid SAS 8708EM2.

Udo

 

[dietmar]

Hi Dietmar,
the same behavior with Megaraid SAS 8708EM2.

But you talk about backula backup, not the new pve vzdump backup?

 

[tom]

I can see this also here.

 

[udo]

But you talk about backula backup, not the new pve vzdump backup?

Right,
I reconice the issue because the VM shut down at the start of the backup-process (with bacula-fd inside the VM).
Like I wrote befoe - it's enough to format an scsi-drive with win2k3 to kill the VM - so it's depends not on bacula.

I would try again to install virtiostor - not the last version (where I got an error, that it's not usable).

Udo