vzdump backup status: backup failed

[Cayuga]

We have about 70 VM's and all of them back up fine, except one. We get the following in the log:

VMID NAME STATUS TIME SIZE FILENAME
168 frll FAILED 00:00:08 command 'lvcreate --size 1024M
--snapshot --name 'vzsnap-coulis.rtr.com-0'
'/dev/pve-nas4/vm-168-disk-1'' failed with exit code 5
TOTAL 00:00:08 0KB


I noticed that this is the only volume with a "real", I'm not sure if it's related, and if it's related, what I should do about it:

dmsetup table | fgrep 168
pve--nas4-vm--168--disk--1: 0 8388608 linear 8:16 1721025024
pve--nas4-vm--168--disk--1-real: 0 4202496 linear 8:16 1813324288

 

[dietmar]

Please try to remove the stale snapshot device with lvremove

 

[Cayuga]

lvscan doesn't show any stale (inactive) snapshots -- is there another way to find them?

 

[dietmar]

lvscan doesn't show any stale (inactive) snapshots -- is there another way to find them?

# dmsetup info

 

[Cayuga]

I have verified using "dmsetup info" that ALL entries are "State: Active". I have also verified that the VM will back up just fine if the VM is not running. But, if the VM is running, I still get the same lvcreate error. I have also verified that migrating the VM to a different cluster node makes no difference.

 

[joelserrano]

I have verified using "dmsetup info" that ALL entries are "State: Active". I have also verified that the VM will back up just fine if the VM is not running. But, if the VM is running, I still get the same lvcreate error. I have also verified that migrating the VM to a different cluster node makes no difference.

Hi,

I know it's a very old post, but just in case I wanted to share my solution.

The VMID in my case: 9002
The storage name in my case: vzpve2

Having very similar scenario, to fix the problem I have done the following:

1- First, as you can see, the 3 disks are marked as "online":

# lvs | grep vzpve2 | grep 9002
  vm-9002-disk-1 vzpve2 -wi-ao   5.01G                                      
  vm-9002-disk-2 vzpve2 -wi-ao   1.00G                                      
  vm-9002-disk-3 vzpve2 -wi-ao   1.00G                                      
#

2- Next: oooops, we have a new one called "-real" (the problematic one).

# dmsetup status | grep 9002
vzpve2-vm--9002--disk--1-real: 0 10502144 linear 
vzpve2-vm--9002--disk--3: 0 2105344 linear 
vzpve2-vm--9002--disk--2: 0 2105344 linear 
vzpve2-vm--9002--disk--1: 0 12591104 linear 
#

3- To remove the stale snapshot disk:

# dmsetup remove vzpve2-vm--9002--disk--1-real

4- Just in case the VM got locked, we unlock it to enable backups again (full/snapshots):

# qm unlock 9002

After that, backups (snapshots) starting working again.

Hope this helps!

Best regards,
Joel.