Hi! Analyzing the problem I found out that the DC found the current CIB corrupted (probably my fault) and used NOT the second-newest CIB, but some "quite old" CIB instead. The CIB used was seven versions behind the current one:
cib: info: get_last_sequence: /var/lib/pacemaker/cib/cib.last was not valid cib: info: cib_file_backup: Archived previous version as /var/lib/pacemaker/cib/cib-0.raw Inspecting cib-0.raw it turns out that it definitely was NOT the previous version. The only thing I had done was (by mistake) to move the location of the CIBs from /var/lib/heartbeat/crm to /var/lib/pacemaker/cib/ while the node was active. As the move was within one filesystem, that really shouldn't have hurt much... Regards, Ulrich Windl >>> "Ulrich Windl" <[email protected]> schrieb am 27.01.2016 um 13:42 in Nachricht <[email protected]>: > Hi! > > I have a question (for SLES11 SP4 with current updates): > Is it a feature that maintenance mode turn off when you do a rcopenais > stop/start? > I had turned maintenance mode on, stopped both nodes of a cluster (rcopenais > stop). Then, when I started the node that was DC before, maintenance mode was > off, and the other node should be fenced. > So I turned on maintenance mode again (with one node active), As soo as I > started the other node, maintenance mode was off again! > > Is this a feature? It's hard to believe! > > Regards, > Ulrich > > > > _______________________________________________ > Users mailing list: [email protected] > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org _______________________________________________ Users mailing list: [email protected] http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
