Hello list! I've recently clustered up my mailserver, using drbd 8.3 with heartbeat 2.1.4 in a moderately simple 2-node R1 configuration.
I can /etc/rc.d/init.d/heartbeat standby or I can pull the power and resources switch over from primary to secondary nodes just lovely. What I am struggling with is a nice clean shutdown of the primary node. I put it on standby and everything runs up fine on the secondary, then I shut down (init 0) the primary. As soon as init runs 'heartbeat stop' on the primary, the secondary closes down perfectly well running resources and I can't seem to persuade it to start them up again. What I observe from the logfiles, is when heartbeat stops on the primary, the secondary runs 'resource start' on all resources, followed immediately by 'resource stop'. This will become jolly inconvenient and not very highly available if I need to close down the primary for any length of time (like to fix the hardware) without simply pulling the plug (which naturally I prefer to avoid). How can I either a: Prevent the secondary node from closing down resources, when primary is cleanly shut down. or b: Persuade the secondary to bring up the resources, even though the primary has been cleanly shut down. Regards Stef _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
