On Thursday 09 December 2010 12:30:09 Andrew Beekhof wrote: > On Thu, Dec 9, 2010 at 12:15 PM, Bart Coninckx <[email protected]> wrote: > > On Wednesday 08 December 2010 12:30:09 Andrew Beekhof wrote: > >> On Mon, Dec 6, 2010 at 10:30 PM, Bart Coninckx > >> <[email protected]> > > > > wrote: > >> > Hi, > >> > > >> > just finished setting up a two-node cluster with > >> > pacemaker-1.0.1-20.3.x86_64 and openais-0.80.3-26.2.x86_64 (OpenSuse > >> > 11.2). > >> > >> 1.0.1??? Please get something newer from clusterlabs.org/rpm > >> > >> > I seem to have quite irradicate (or so I persieve it) behaviour from > >> > the nodes where they go both offline, sometimes I get to have one > >> > online and than sometimes (until reboots) I have one offline in a > >> > pending state. > >> > > >> > I think I remedied the two nodes offline situation at one point by > >> > enabling stonith again and defining a stonith resource, but later I > >> > had one of the above situations again. It seems hard to get both > >> > nodes running and doing failovers. > >> > > >> > Where is a good place to start looking? > >> > > >> > I enabled debugging on some daemons and doing this on crmd showed me: > >> > > >> > ERROR: send_ais_message: Not connected to AIS > >> > > >> > > >> > Maybe that's a hint. All daemons are running though. > >> > > >> > Thx!! > >> > > >> > Bart > >> > _______________________________________________ > >> > Linux-HA mailing list > >> > [email protected] > >> > http://lists.linux-ha.org/mailman/listinfo/linux-ha > >> > See also: http://linux-ha.org/ReportingProblems > >> > >> _______________________________________________ > >> Linux-HA mailing list > >> [email protected] > >> http://lists.linux-ha.org/mailman/listinfo/linux-ha > >> See also: http://linux-ha.org/ReportingProblems > > > > Hi Andrew, > > > > thank you for getting back to me. We like to stay in the realm of the > > offered distro RPMs for ease of upgrade, but I do have an option to go > > to opensuse 11.3 and have pacemaker-1.1.2.1-2.1.1.x86_64, > > openais-1.1.2-2.1.1.x86_64 and corosync-1.2.1-1.2.x86_64. > > . > > > > Would that be a better option? > > From the pacemaker side, absolutely. > > > These machines are Xen hypervisors and that > > would force us to go to Xen 4.0.0 on opensuse 11.3, which is not quite > > stable. > > > > One other option is to move to SLES 11 and the HAE extension. They have: > > pacemaker-1.1.2-0.2.1.x86_64.rpm > > openais-1.1.2-0.5.19.x86_64.rpm > > corosync-1.2.1-0.5.1.x86_64.rpm > > > > Do you have any recommondation in this? > > Also a very good option. They funded most of the pacemaker > development, so they know a thing or two about putting together a > solid stack.
Happy to report that things go a lot smoother on SLES11SP1 with the HAE extension. Migration of the Xen guest resource does not appear to be a cold migration up till now, but I'll try to figure that out and if I don't succeed, come back with a separate posting. cheers, Bart _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
