Cluster backup problem

C

crewze

New Member
Oct 17, 2009
24
1
1
I have a cluster of 2 nodes that is working fine (proxmox and proxmox1). I have setup backups for the master node (proxmox) to an external NFS and that is working fine.

I setup the backup for VMs on the other node (proxmox1) to the same NFS and they worked for a while and then stopped working. I deleted them and recreated them and they are still not working. When I look in the logs for the backup I see:

Dec 22 04:00:01 promox /USR/SBIN/CRON[20726]: (root) CMD (vzdump --quiet --node 2 --snapshot --storage Tuesday --mailto george@shafer.ca 1008^I1015^I1016)

The VMs I am trying to backup are 1008, 1015 and 1016. When I try running this from the command line it comes back after a few seconds and does not back anything up or send me an email.

This is the log from the backup that is working fine:

Dec 22 03:00:01 promox /USR/SBIN/CRON[8605]: (root) CMD (vzdump --quiet --node 1 --snapshot --compress --storage Tuesday --mailto george@shafer.ca 1009)

Not sure what to try next. Any ideas?
 
Last edited:
crewze said:
The VMs I am trying to backup are 1008, 1015 and 1016. When I try running this from the command line it comes back after a few seconds and does not back anything up or send me an email.
Click to expand...

Maybe the job is for the wrong node (--node X)? (You moved the VMs?)
 
I created the command from the Proxmox admin panel so it is all selected by drop down boxes. I have not moved any of the VMs.
 
When I run this command from the command line:

vzdump --quiet --node 2 --snapshot --storage Tuesday --mailto george@shafer.ca 1008

It goes away for a few seconds and then returns to the command prompt. It does not back anything up or send me an email.
 
crewze said:
When I run this command from the command line:

vzdump --quiet --node 2 --snapshot --storage Tuesday --mailto george@shafer.ca 1008

It goes away for a few seconds and then returns to the command prompt. It does not back anything up or send me an email.
Click to expand...

Any you run that on node 2 ?
 
No --- I ran it on node 1.

I was under the impression that all management for the cluster was done on the master node. Do you not configure all of the backups on the master?
 
crewze said:
I was under the impression that all management for the cluster was done on the master node. Do you not configure all of the backups on the master?
Click to expand...

Sure - but we want to test now what happens if we run manually. So please start on the correct node, or start without --node X option.
 
OK. I ran it on the second node:

proxmox1:~# vzdump --snapshot --storage Tuesday --mailto george@shafer.ca 1008mkdir /mnt/pve/Tuesday: File exists at /usr/share/perl5/PVE/Storage.pm line 2104


If I run it on the main node I get:

promox:/var/log# vzdump --snapshot --storage Tuesday --mailto george@shafer.ca 1008
INFO: starting new backup job: vzdump --snapshot --storage Tuesday --mailto george@shafer.ca 1008
ERROR: Backup of VM 1008 failed - unable to find VM '1008'
INFO: Backup job finished with errors


Anything else I can run to help figure this out?
 
crewze said:
OK. I ran it on the second node:

proxmox1:~# vzdump --snapshot --storage Tuesday --mailto george@shafer.ca 1008mkdir /mnt/pve/Tuesday: File exists at /usr/share/perl5/PVE/Storage.pm line 2104
Click to expand...

Strange, I thought I already fixed that bug - what version do you run? What is the output of

# pveversion -v
 
promox:~# pveversion -v
pve-manager: 1.4-10 (pve-manager/1.4/4403)
qemu-server: 1.1-8
pve-kernel: 2.6.24-16
pve-qemu-kvm: 0.11.0-2
pve-firmware: 1
vncterm: 0.9-2
vzctl: 3.0.23-1pve3
vzdump: 1.2-5
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1
 
I also need the storage configuration - please can you post the contents of /etc/pve/storage.cfg ?
 
Last edited:
Please check logs for errors and let us know what you see? Did you try different browser?
 
I have the same problem.
When I run
Code: 
vzdump --node 2 --snapshot --storage NFS-246 --mailto it3@slata.ru 252
on cluster node, I don`t see any messages.
When I run
Code: 
vzdump --snapshot --storage NFS-246 --mailto it3@slata.ru 252
on cluster node, I see that
Code: 
INFO: starting new backup job: vzdump --snapshot --storage NFS-246 --mailto it3@slata.ru 252
ERROR: Backup of VM 252 failed - unable to find VM '252'
INFO: Backup job finished with errors
And when I run
Code: 
vzdump --snapshot --storage NFS-246 --mailto it3@slata.ru 252
on second node, I see that
Code: 
mount.nfs: 10.0.4.246:/smpool/backups failed, reason given by server: Permission denied
command '/bin/mount -t nfs 10.0.4.246:/smpool/backups /mnt/pve/NFS-246' failed with exit code 32
 
Decard said:
I have the same problem.
When I run
Code: 
vzdump --node 2 --snapshot --storage NFS-246 --mailto it3@slata.ru 252
on cluster node, I don`t see any messages.
Click to expand...

Any what is the problem? When you use '--node 2' the job only runs when you are on node 2 - that is the purpose of that option.
 
I think the main problem is in misunderstanding the backup function. The file /etc/cron.d/vzdump will also synced between all cluster nodes. The option "--node X" allows running only on "Node X". All other nodes be ignored. Take a look in /etc/cron.d/vzdump on all your nodes (incl. master node).
A further problem is, here in my installation, that the mount points for the backup storage defined in /etc/pve/storage.cfg won't be created on the nodes. So the backups have failed in the past. If you create the directories manually, there is no problem with backing up your VMs.
 
You must log in or register to reply here.
Top Bottom