state: nosync...?

[Robstarusa]

What causes this? I moved all my vm's from vm2 to vm and then made vm master and then shut down vm2 to upgrade it from 4G to 8G ram. I then restarted it.....vm is still the master The status in the main proxmox ve page shows:

Hostname IP Address Role State Uptime Load CPU IODelay Memory Disk
vm 192.168.75.16 Master active 1 day 01:13 2.17 65% 22% 25% 20%
vm2 192.168.75.29 Node nosync 02:03 3.09 24% 44% 11% 5%

I can still migrate machines back & forth and it seems to work ok... Any suggestions appreciated. Blinking red is kind of annoying Also, I'm thinking of adding a 3rd node. Is/are there any limitation(s) I should be aware of when adding the 3rd node? It is a be-2350 (2.1ghz dualcore) with 4G ram. Do I have to have an even number of nodes?

 

[dietmar]

Ins't there any hint in syslog why the node does not sync?

Anyways, stop the mirror and tunnel service, then delete /etc/pve/cluster.conf and simply join the node again. Does that help?

There is no limit on the nember of nodes.

- Dietmar

 

[Robstarusa]

I just re-checked the status now and it is normal. ARGH. If it happens again I'll check syslog.

 

[dietmar]

The cluster is synced every 3 minutes (not immediately).

- Dietmar

 

[Robstarusa]

I had the message for a couple of hours :-(

 

[burtsevyg]

I see this message, when restore my vm to drbd storage. I think it's normal.

 

[bougui]

on the non master you can run

pveca -s

and for me it was a typo error in the /root/bashrc file on the master that was giving and exit code 2 to rsync

I correct the error in my bash rc file and voila ! the cluster was fine after a

pveca -s

YMMV,

Guillaume

 

[musicman3569]

In case anyone else runs into this problem, it seems that errors with storage can cause a lot of seemingly unrelated issues in Proxmox.

I had this exact problem, and in my case I had an added backup storage (via NFS) to a FreeNAS computer that had frozen. Once I got the FreeNAS booted, the sync went back to normal and I could stay logged into the web interface.

The real clue was in /var/log/syslog:

May 23 11:44:25 virtsrv1 kernel: nfs: server 10.10.1.9 not responding, still trying
May 23 11:44:35 virtsrv1 pvemirror[6607]: starting cluster syncronization
May 23 11:44:45 virtsrv1 pvemirror[6607]: syncing vzlist from '10.10.1.2' failed: 500 read timeout

Short story: wierd issues -> check any network storage

Hope it helps!