Corrupted Filesystem after reinstall of Virtualmachine

[timetraxx]

Hi All, yesterday evening, I reinstall with our Customer different VM's on our 4 Node Proxmox Cluster.
After 6h our new VM's became corrupt ext4 Filesystems. We used the IDE HD Driver instead of the VIRTIO dirver.
I run a lot fsck aessions on the VM's - now way to repair the filesystem. we reinstall the VM's again but now with the VIRTIO driver for our Didks.
It is possible that IDE drivers are not a good Idea to virtualize systems on Proxmox?

Following entries in /var/log/messages EXT4-fs error (device dm-0): ext4_lookup: deleted inode referenced: 2093069
EXT4-fs error (device dm-0): ext4_lookup: deleted inode referenced: 2093069
EXT4-fs error (device dm-0): ext4_lookup: deleted inode referenced: 2093069
EXT4-fs error (device dm-0): ext4_lookup: deleted inode referenced: 2093069
EXT4-fs error (device dm-0): ext4_lookup: deleted inode referenced: 2093069

My Actual Setup:

proxmox-ve-2.6.32: 3.2-121 (running kernel: 2.6.32-27-pve)
pve-manager: 3.2-1 (running version: 3.2-1/1933730b)
pve-kernel-2.6.32-27-pve: 2.6.32-121
pve-kernel-2.6.32-23-pve: 2.6.32-109
lvm2: 2.02.98-pve4
clvm: 2.02.98-pve4
corosync-pve: 1.4.5-1
openais-pve: 1.1.4-3
libqb0: 0.11.1-2
redhat-cluster-pve: 3.2.0-2
resource-agents-pve: 3.9.2-4
fence-agents-pve: 4.0.5-1
pve-cluster: 3.0-12
qemu-server: 3.1-15
pve-firmware: 1.1-2
libpve-common-perl: 3.0-14
libpve-access-control: 3.0-11
libpve-storage-perl: 3.0-19
pve-libspice-server1: 0.12.4-3
vncterm: 1.1-6
vzctl: 4.0-1pve4
vzprocps: 2.0.11-2
vzquota: 3.1-2
pve-qemu-kvm: 1.7-4
ksm-control-daemon: 1.1-1
glusterfs-client: 3.4.2-1

Storage MSA 2000, fibrechannel Multipath ( All 4 Nodes are connected to the MSA)

OS Type VM:
RHEL 6.5

 

[udo]

Hi,
hard to read your input. If you connect via https to the forum (which is an good idea) use the standard editor (Settings -> Gereral Settings).

For an linux is IDE and virtio very stable. I assume it's has someting to do with your underlaying storage (like an ext-volume, which is mounted on more than one node together).

Udo

 

[timetraxx]

Hi Udo,

thnx for Your editor help

we have connected al 4 Nodes to the Storage with fibrechannel multipath.
We add the LVM (volume group) under Datacenter with the following options:

Nodes: All
Enable: true
Shared:True

Do you mean that the Nodes All Options can be a problem?

Regards

Martin

 

[udo]

Hi Udo,

thnx for Your editor help

we have connected al 4 Nodes to the Storage with fibrechannel multipath.
We add the LVM (volume group) under Datacenter with the following options:

Nodes: All
Enable: true
Shared:True

Do you mean that the Nodes All Options can be a problem?

Regards

Martin

Hi Martin,
no "Nodes All" is the right config (if all nodes has access to the storage).

Strange, I use lvm on different storages without trouble (drbd, FC) but without multipath.
Perhaps, there is an problem (do you mask the single devices in lvm.conf?).

Can you try with single-path on one node?

Udo

 

[timetraxx]

Hi Udo,

yes the device are masked.
I've no chance to switch off from multi to singlemode. But I'll take this point and check in our LAB to test.

I myke some checks but I can see nothing which can be a problem ...

-> VoDo

 

[udo]

Hi Udo,

yes the device are masked.
I've no chance to switch off from multi to singlemode. But I'll take this point and check in our LAB to test.

I myke some checks but I can see nothing which can be a problem ...

-> VoDo

Hi,
one thing left to test: can you switch the cache-settings of the vm-disk to writethrough?! (power cycle the VM).
Like for lvm-storage with blocksize > 512 it's neccesary.

Udo