It can happen that a Oracle VM Server somehow doesn’t want to leave a server pool. I encountered this when I removed a clustered server pool (by, of course, first de-assigning the server nodes assigned to it, and then removing the server pool). It so happened to be that two servers in that server pool were down during the change because of the distributed power management policy. By the time these servers came online, the server pool wasn’t there, the NFS share wasn’t either, and both servers were in a non-existent server pool which I couldn’t remove or alter.
In such a situation you’d want to reset an Oracle VM Server. What helped me were these excellent steps posted on the Oracle Forums by user “627349″, these work like a charm:
#Stop ovs-agent service ovs-agent stop #Delete /etc/ovs-agent/db on the server(s) rm /etc/ovs-agent/db #Empty the /etc/ocfs2/cluster.conf /dev/null> /etc/ocfs2/cluster.conf #Start the ovs-agent service ovs-agent start
I always reboot the Oracle VM Server afterwards, just to be sure. The Oracle VM Server will then be an unassigned server in Oracle VM Manager, so you can assign it again to some server pool.Reset an Oracle VM Server,