[Linux-cluster] clusvcadm

[Digimer]

On 07/05/14 06:41 PM, Paras pradhan wrote:
> Yes the fencing failed because of the network issues and I had to reboot
> one of the manually and it came back normal. But now I am seeing this
> DLM is blocked.

Then this is why you're stuck. RGmanager, gfs2 and clvmd all use DLM and 
they will block when DLM blocks. When a fence is called, fenced informs 
dlm, and dlm remains blocked until fenced tells dlm that it's safe to 
recover.

If you are *certain* that the one nodes is powered off (or rebooted and 
cman/rgmanager was not started), you can run 'fence_ack_manual' to tell 
fenced that the node was fenced. Be very careful with this; improper use 
can corrupt your cluster.

Alternatively, restart all nodes.

-- 
Digimer
Papers and Projects: https://alteeve.ca/w/
What if the cure for cancer is trapped in the mind of a person without 
access to education?