Proxmox Cluster and HA

[Zriouil]

Hello,

Please I need help

I created a cluster Proxmox of two nodes with DRBD storage, I configured Fencing to have the HA.
I tested the high availability by stopping the rgmanager of a node, the virtual machines are switched to the other.
So everything was fine

After a restart of servers, the cluster is no more.

when I put pvecm nodes, this is what i have

Node Sts Inc Joined Name
1 M 76 2015-04-23 09:49:24 ahmed
2 X 0 zr

In the second node, i have cman_tool: cannot open connection to cman, is it running ?
and i can't start cman anymore:
Cannot find node name in cluster.conf
Unable to get the configuration
Cannot find node in cluster.conf
cman_tool: corosync daemon didn't start Check cluster logs for details
[FAILED]

about the web interface, in the first URL I have the 2 nodes but in the other one only

Thank in advance.

 

[wolfgang]

Hi,
you can try to copy the config from the other node, but I'm not sure if this helps!

/etc/cluster/cluster.conf

 

[Zriouil]

Hi, Thank for reply

I can't modify cluster.conf
even the change of access rights is not possible

 

[wolfgang]

Do you are talking about /etc/pve/cluster/cluster.conf or /etc/cluster/cluster.conf ?

 

[Zriouil]

I talk about /etc/pve/cluster.conf !

 

[wolfgang]

The /etc/pve/* is set read only because you have probably no quorum!
you can check this with
pvecm status

Can you send the cluster.conf?

 

[Zriouil]

thank you very much dear
the problem is solved
the file /etc/cluster/cluster.conf is editable

 

[Zriouil]

yes, i didn't have the quorum.

<?xml version="1.0"?>
<cluster config_version="6" name="mycluster">
<cman two_node="1" expected_votes="1" keyfile="/var/lib/pve-cluster/corosync.authkey"/>
<fencedevices>
<fencedevice agent="fence_ilo" ipaddr="192.168.1.200" login="root" name="fenceAHMED" passwd="123456"/>
<fencedevice agent="fence_ilo" ipaddr="192.168.1.210" login="root" name="fenceZRI" passwd="123456"/>
</fencedevices>
<clusternodes>
<clusternode name="ahmed" nodeid="1" votes="1">
<fence>
<method name="1">
<device action="reboot" name="fenceAHMED"/>
</method>
</fence>
</clusternode>
<clusternode name="zriouil" nodeid="2" votes="1">
<fence>
<method name="1">
<device action="reboot" name="fenceZRI"/>
</method>
</fence>
</clusternode>
</clusternodes>
<rm>
<pvevm autostart="1" vmid="100"/>
<pvevm autostart="1" vmid="101"/>
</rm>
</cluster>
~


the problem is already resolver but I have a problem when I turn off the server, it loses quorum. suddenly I have to re-edit the file cluster.conf every time