2007/7/7, Ciro Iriarte <[EMAIL PROTECTED]>:
Hi, just arrived to the office to continue with my tests and found that my cluster is in a split brain condition. The VLAN used for "cluster interconnect" had intermitent troubles for almost 3 hours :sObviously, the two nodes wanted to become primary. The weird think is, after the link was restablished, both were still displaying the other one as OFFLINE. Isn't hb supposed to find the other node again?. Trying to shutdown one of the nodes didn't work, the script never returned the prompt, so i had to kill the heartbeat process and restarting just one of them took to cluster to normal state again. View from node1: ============ Last updated: Fri Jul 6 22:17:36 2007 Current DC: asusis-ope1 (34b56337-2402-4e2d-a66a-f0f0c2cee1b5) 2 Nodes configured. 3 Resources configured. ============ Node: asusis-ope1 (34b56337-2402-4e2d-a66a-f0f0c2cee1b5): online Node: asusis-ope2 (9b4731c8-163c-482a-93f4-084a5067a414): OFFLINE resource_ip_10_129_4_236 (heartbeat::ocf:IPaddr2): Started asusis-ope1 Master/Slave Set: ms-hadata0 hadata0:0 (heartbeat::ocf:drbd): Master asusis-ope1 hadata0:1 (heartbeat::ocf:drbd): Stopped fs_hadata (heartbeat::ocf:Filesystem): Started asusis-ope1 View from node2: ============ Last updated: Fri Jul 6 22:17:53 2007 Current DC: asusis-ope1 (34b56337-2402-4e2d-a66a-f0f0c2cee1b5) 2 Nodes configured. 3 Resources configured. ============ Node: asusis-ope1 (34b56337-2402-4e2d-a66a-f0f0c2cee1b5): OFFLINE Node: asusis-ope2 (9b4731c8-163c-482a-93f4-084a5067a414): online resource_ip_10_129_4_236 (heartbeat::ocf:IPaddr2): Started asusis-ope2 Master/Slave Set: ms-hadata0 hadata0:0 (heartbeat::ocf:drbd): Stopped hadata0:1 (heartbeat::ocf:drbd): Started asusis-ope2 Ciro
Oops, forgot about the version, it's from the Build Service asusis-ope1:~ # rpm -q heartbeat heartbeat-2.1.0-28.1 Ciro _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
