My expectation with an active/active cluster is that the resources are running on both servers at the same time and they are accessible from either server simultaneously. If either server goes down, the collocated IP address should mean that all the traffic automatically flows to the one server that is still up.
I am quite confused as to why there is a "process" of switching from one node to the other. I am starting to wonder if active/active truly means active/active. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Ulrich Windl Sent: Thursday, August 02, 2012 5:08 AM To: [email protected] Subject: [Linux-HA] Antw: Failover Failure Hi! Things I've seen with SLES11 SP1 and a two-node cluster was this: If only one node is alive, it takes quite some time (about 20 minutes) until the node becomed new domain controller. Ocassionally even if the second node comes up, both nodes cannot agree who's DC for quite some while. Nothing will happen unless there's a DC, and nothing will happen if the cluster has no quorum. Maybe check those. Reading syslog messages may help you to understand what's going on. I still don't understand all the stuff thst's happening, but after some time you learn what "just happens". BTW: Is there anybody there to explain all the "synapse" and "firing" stuff? Years ago I was quite good in neurophysiology, but I cannot relate it to cluster resources ;-) Regards, Ulrich >>> "Yount, William D" <[email protected]> schrieb am >>> 02.08.2012 um 10:43 in Nachricht <11893698c5a8c84ea0b5484deb54e375025a983...@dcxprcl018.cnf.prod.cnf.com>: > Attached is my cib.xml file. > > I have a two node DRBD cluster setup in Active/Active. For whatever > reason, it seems all my resources are attached to Node2. What I mean > by that is that although the resources show that they are collocated, > whenever I turn Node2 off or unplug a cable from Node2, then the > cluster goes down. I wait to see if they come back up on the other > node (although they should already be running as it is an > Active/Active cluster) but they never do, even after 10 minutes. With Node2 > off, I can't even ping the collocated IP address. > However, if I turn off Node1 while Node2 is running, nothing goes down. > > I am using the LCMC to give me a graphical overview of the setup and > the screen seems to indicate that everything is okay. I believe it has > to do with my fencing agent which is pacemaker. I know that even > though it is set to turn a node off if there is an issue, the node > never seems to shutdown. It complains that devices are busy and it can't > reboot. > > I am just hoping someone can take a look at my configuration and see > if there is anything that stands out. If it is the fencing agent, is > there a better fencing agent? > > > William > > _______________________________________________ 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
