Greetings, having posted a question earlier about how to monitor my networkport once it is down, I now have three monitored resources which when failing are all tried restarted on the same node as they initially were started on (this is ok for the two actual services, but when a cable is broken, restarting does no good).
My solution lies within setting up scores and relying on the failure count to acually failover all resources within a group once the score becomes lower than my other node. My setup is: - Two nodes, installed with sles10 sp2, running heartbeat 2.1.3 (EVERYTHING setup through hb_gui) - One group which contains six resources (three of these are monitored) - The group has resource_stickiness="100" and failure_resource_stickiness="-300" (resources have none) - Two location constraints, both have the group as resource. - Location constraint "A" has a score of 200 and #uname equals server node "A" - Location constraint "B" has a score of 100 and #uname equals server node "B" After i reboot both servernodes and "A" takes the resources i try to run showscores.sh on "A" but only the resources are displayed as "resource.servernode" - no score, no failurecount, no nothing? I then introduce an error on one of the monitored services and after two errorcounts the group is failed over to servernode "B". I run showscores.sh on "A" again but I receive the same answer again? I now introduce an error on servernode "B" and after two errorcounts, resources are failed over to "A". Trying to fail "A" once again just results in the service trying to become restarted on "A" from now on - it is like I have used up all my number of failover tries? How am I supposed to maintain 100% automatic failover? How can I check my scores? Thanks in advance! -- View this message in context: http://www.nabble.com/showscores.sh-does-not-work-and-failover-is-unpredictable--tp20722584p20722584.html Sent from the Linux-HA mailing list archive at Nabble.com. _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
