[Linux-cluster] problem with virbro, virbr-nic

[Andrew Kerber]

I have gfs2 shared file system configured and I am using fench-virsh as the
fencing system on RHEL 7.  On virtuals.  One of our technicians made a
mistake and destroyed on of the VM's, which re-created by cloning the
second node of the cluster.  I reconfigured everything by removing the
cluster configuration using pcs cluster destroy, uninstalling lib virt and
dependent software, and reinstalling.

However, the virbr0 and virbro-nic interfaces come up every time with the
same mac address on both nodes.  I am pretty confident this is the cause of
the problem we are having, in that the gfs file system will not mount
through the cluster software (pcs).  The file system always comes up in
status failed, and will not clear.  Is there a configuration file somewhere
that needs to be manually deleted when reinstalling libvirt so that a new
mac address is generated?

-- 
Andrew W. Kerber

'If at first you dont succeed, dont take up skydiving.'
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/linux-cluster/attachments/20160912/cd8bc42a/attachment.htm>