Hi, On Mon, Aug 23, 2010 at 12:52:38PM -0500, [email protected] wrote: > I have two older desktop pcs that I have set up with drbd, corosync and > pacemaker > No problems there > My hat is off to y'all > Great software > I have set up the colocation rules so that all apps run on one machine or the > other > If I break the communication link between them I get split-brain > So I figure stonith suicide is my only option
suicide is not an option at all. You need a proper fencing device. And you need redundant communication links. Thanks, Dejan > When the communication goes down the machine hosting the apps > continues to function properly > The other machine sort of "hangs" > > > ####################################################################### > With good communication link > > [r...@abctest1 ~]# crm_mon -1 > ============ > Last updated: Fri Aug 20 16:48:13 2010 > Stack: openais > Current DC: abctest1 - partition with quorum > Version: 1.0.9-89bd754939df5150de7cd76835f98fe90851b677 > 2 Nodes configured, 2 expected votes > 8 Resources configured. > ============ > > Online: [ abctest1 abctest2 ] > > Master/Slave Set: ms-drbd0 > Masters: [ abctest2 ] > Slaves: [ abctest1 ] > fs0 (ocf::heartbeat:Filesystem): Started abctest2 > apache (ocf::heartbeat:apache): Started abctest2 > mysql (ocf::heartbeat:mysql): Started abctest2 > tomcat (ocf::heartbeat:tomcat): Started abctest2 > clusterip1 (ocf::heartbeat:IPaddr2): Started abctest2 > clusterip2 (ocf::heartbeat:IPaddr2): Started abctest2 > Clone Set: fencing > Started: [ abctest2 abctest1 ] > > > > [r...@abctest2 ~]# crm_mon -1 > ============ > Last updated: Mon Aug 23 12:36:56 2010 > Stack: openais > Current DC: abctest1 - partition with quorum > Version: 1.0.9-89bd754939df5150de7cd76835f98fe90851b677 > 2 Nodes configured, 2 expected votes > 8 Resources configured. > ============ > > Online: [ abctest1 abctest2 ] > > Master/Slave Set: ms-drbd0 > Masters: [ abctest2 ] > Slaves: [ abctest1 ] > fs0 (ocf::heartbeat:Filesystem): Started abctest2 > apache (ocf::heartbeat:apache): Started abctest2 > mysql (ocf::heartbeat:mysql): Started abctest2 > tomcat (ocf::heartbeat:tomcat): Started abctest2 > clusterip1 (ocf::heartbeat:IPaddr2): Started abctest2 > clusterip2 (ocf::heartbeat:IPaddr2): Started abctest2 > Clone Set: fencing > Started: [ abctest2 abctest1 ] > > > ####################################################################### > After links goes down > > [r...@abctest1 ~]# crm_mon -1 > ============ > Last updated: Mon Aug 23 12:37:37 2010 > Stack: openais > Current DC: abctest1 - partition with quorum > Version: 1.0.9-89bd754939df5150de7cd76835f98fe90851b677 > 2 Nodes configured, 1 expected votes > 8 Resources configured. > ============ > > Node abctest2: UNCLEAN (offline) > Online: [ abctest1 ] > > Master/Slave Set: ms-drbd0 > Slaves: [ abctest1 ] > Stopped: [ drbd0:0 ] > Clone Set: fencing > Started: [ abctest1 ] > Stopped: [ stonith-suicide:0 ] > > > [r...@abctest2 ~]# crm_mon -1 > ============ > Last updated: Mon Aug 23 12:37:57 2010 > Stack: openais > Current DC: abctest2 - partition WITHOUT quorum > Version: 1.0.9-89bd754939df5150de7cd76835f98fe90851b677 > 2 Nodes configured, 2 expected votes > 8 Resources configured. > ============ > > Node abctest1: UNCLEAN (offline) > Online: [ abctest2 ] > > Master/Slave Set: ms-drbd0 > Masters: [ abctest2 ] > Stopped: [ drbd0:1 ] > fs0 (ocf::heartbeat:Filesystem): Started abctest2 > apache (ocf::heartbeat:apache): Started abctest2 > mysql (ocf::heartbeat:mysql): Started abctest2 > tomcat (ocf::heartbeat:tomcat): Started abctest2 > clusterip1 (ocf::heartbeat:IPaddr2): Started abctest2 > clusterip2 (ocf::heartbeat:IPaddr2): Started abctest2 > Clone Set: fencing > Started: [ abctest2 ] > Stopped: [ stonith-suicide:1 ] > > ####################################################################### > At this point abctest1 is "hung" > The machine still functions but corosync will not shutdown nor will is > reestablish communication after the link is is connected > A pkill -9 cocosync is required to bring it down > Is the how suicide is supposed to behave > > Richard > _______________________________________________ > 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
