teigland at redhat.com
Mon Aug 20 20:16:15 UTC 2007
On Mon, Aug 20, 2007 at 04:12:32PM -0400, Patricio A. Bruna wrote:
> This is RHEL 5.0.
> Im using the original cman init script provided by: cman-2.0.60-1.el5
You'll want to update that; the cluster will be clobbered by the xen init
script without the fix. You can work around that for now by doing the
'chkconfig cman off' and then starting the cluster manually on both nodes
with 'service cman start'.
I can't say that the init script is your main or only problem here, but we
need to eliminate that first.
> ----- "David Teigland" <teigland at redhat.com> escribi??:
> > On Mon, Aug 20, 2007 at 02:21:45PM -0400, Patricio A. Bruna wrote:
> > > Here goes the information.
> > Is this RHEL 5.0? Are you using the original cman init script or the
> > updated one? Could you 'chkconfig cman off' on both nodes, then
> > reboot
> > both nodes, when both have come back, then manually 'service cman
> > start'
> > on both nodes. If they get stuck could you collect from both nodes:
> > cman_tool status
> > cman_tool nodes
> > group_tool -v
> > group_tool dump > groupd.txt
> > group_tool dump fence > fenced.txt
> > /var/log/messages (anything related to aisexec, cman, groupd, fencing,
> > ...)
> > Thanks,
> > Dave
More information about the Linux-cluster