Brett Maton <brett.maton@...> writes: > > Hi Takatoshi, > > I probably did start the slave node manually. > I think I'm getting stuck with not knowing how to properly manage start-up > and shutdowns with the cluster. > I can get it all up and running (somehow ;)), but I don't understand how > to restore the master and slave to their original states. > > Taking a new base backup / rsync'ing the WAL logs is no problem, just > how-to manage the service with crm is tripping me up and I haven't figured > out how to read the logs (/var/log/corosync.log) yet.... > > Again, thank you all for your patience. > Brett
In fact, the sample configuration given by Takatoshi takes use of ping service to state if a node replies (as I can understand it..) I choose another way to make its great agent drive well my two nodes with Slave and Master combination : First I've set up a DRBD resource, because for other reason I need a share storage. Then I have a group of Services that follows the DRBD Master (which is a MS clone in fact) Finally my PGSQL resource is a MS (MultiState) too, that follows the DRBD as well (thanks to colocation rules). So the PG Master MUST always be same as DRBD Master, and this is moreover fine because my Services contains a VIP with which I can connect to the Master PG (in read/write). Additionnally, I can use the PG Slave as a read/only database. Got the scheme ? _______________________________________________ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
