[SOLVED] Problem permission /etc/pve/nodes

[merrow]

Hi,

This morning the permission have changed for the /etc/pve/nodes.
The directory is read and execute for root and www-data but it isn't read, write, execute for root.
So my backup for my VM is failed.
I don't know what happened and It's not possible for me of change permission.

I have test with pmxcfs -l and change the permission

But after I don't stop pve-cluster and I have this message if restart cluster
crit: unable to aquire pmxcfs lock: Resource temporarily unavailable


My version

pveversion
pve-manager/4.0-48/0d8559d0 (running kernel: 4.2.2-1-pve)

Thanks for your help

Best regards

 

[dietmar]

Maybe you run a cluster and lost quorum?

 

[merrow]

I removed one of my node cluster, for the reinstall it three days ago

My proxmox cluster has only one node

How can I do to restore rights ?

thanks

 

[t.lamprecht]

So did you executed pvecem delnode <oldnode> after you removed the node?

If not gain temporary quorum with

pvecm expected 1
pvecm delnode old-node-name

If you have done this and it does not work please post output from

pvecm status

 

[merrow]

pvecm status

------------------
Date:             Wed Apr 20 12:00:59 2016
Quorum provider:  corosync_votequorum
Nodes:            1
Node ID:          0x00000001
Ring ID:          60
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   1
Highest expected: 1
Total votes:      1
Quorum:           1 
Flags:            Quorate

Membership information
----------------------
    Nodeid      Votes Name
0x00000001          1 ip_my_server (local)

I have restart my server and the problem is solved, but in my directory /etc/pve/nodes.
I look my old node, this normal ?

Thanks for your help

 

[t.lamprecht]

I have restart my server and the problem is solved, but in my directory /etc/pve/nodes.
I look my old node, this normal ?

Yes this is by design, so that no config file from the old node gets lost. If you are sure that you have saved all configuration and you do not need the old node anymore you can simply delete it.

When deleting a node we do not touch its data, we only remove it from the cluster configuration.

Your status output seems all good .

 

[Jorge Fernando]

In my proxmox 3.4 installation I have a similar problem. I can not write to the /etc/pve directory.

root@proxmox26:/etc/pve# pveversion
pve-manager/3.4-1/3f2d890e (running kernel: 2.6.32-37-pve)

root@proxmox26:/etc/pve# pvecm status
cman_tool: Cannot open connection to cman, is it running ?

Please Help!!!

 

[t.lamprecht]

In my proxmox 3.4 installation

Inevitable note that you should upgrade as PVE 3.4 is EOL, but as this won't help you her lets try to investigate it:

First I would like to get a little more detail about your setup, please. E.g., how many nodes, is it a cluster or not (just to be sure) and since when happens this.
Could you try to restart cman and eventually pve-cluster?

service cman restart
service pve-cluster restart

Please answer in your stand alone thread here:
https://forum.proxmox.com/threads/problem-permission-etc-pve-nodes.37634/
(I just saw now that you opened also a stand alone thread)