Cluster Node Login loop

[kickinz1]

Hi eveyone!

I'm very please with this version of proxmox which looks promising! Good Job all!

I'm currently running a small 4 systems cluster to test proxmox v2.0

I've been able to add all nodes to the cluster, but if it is ok in the WebUi to see all parameters of node 1 and node 2, for nodes 3 & 4, the WebUI is always asking me to login, and drop 'data load errors'. It loop quite fast so it's quite annoying (1-2s).

I can access each nodes WebUI without a trouble. I can ssh them, etc...

Any Idea?

PS: I've been looking in the forum, and it seems I'm the only one, I've already reloade all 4 systems....

Thanks for any help!

 

[tom]

make sure that all nodes are running the latest version and check if all 'cman' service is running on all nodes.

 

[dietmar]

Is the time correct on all nodes?

 

[kickinz1]

Hi thanks for your answers,

I've enabled cman service at boot time update-rc.d cma enable because it wasn't started.
On reboot cman is stopped (fenced) on nodes 3&4, forcing it to start doesn't solve anything.

I've checked the version : all nodes have been updated to beta3 (2.0/784729f4)

BUT I have a different login interface between nodes 1&2 and nodes 3&4, On node 1&2 I have 4 text zones (user name, password, authentication, language), on nodes 3&4 only three (user name, password, authentication).

I have done the updates at the same time on the 4 nodes.. What could it be?

I'm checking further, but I'd be pleased if you can give any advice...

they wer'nt at the same date, now it's done, but still no luck, the login panel is different so it doesn't seems to be the same version.... quite strange repositories are the same, Installation media was the same, Upgrade was done the 4 at the same time...

 

[dietmar]

What is the output of

# pveversion -v

on node 3 (or 4).

 

[Fleg]

I had the same problem. I solved it by setting the same time on all nodes.