Greetings, First time posting to this list; not sure if I should post the results of hb_report (it's quite big so I'll leave it off).
I have 4 nodes each of which is running these resource agents, in this order: group_01: IPAddr Filesystem_01 Filesystem_NN gluster_instance ... group_NN: ... (I'm trying to produce a HA filesystem using gluster) Everything works until I failover or standby one of the nodes. At that point, everything migrates properly except now some of the nodes can't ping some of the newly moved IP addresses. When I run 'arp -n' and compare the mac addresses I see that the old mac address is not being updated. I would think that I could use send_arp to tell the machine the correct address but it seems to have no effect? I've tried to 'arp -d <old_ip>' which seems to wipe part of the information but then it comes back. On each node I have 4 NICs, one of which has a private, non-routable ip address, the remaining 3 are bonded using link aggregation bonding. This seems to be pretty fundamental functionality for heartbeat to be having trouble with. Is there something obvious that someone could point me to? Thanks much, JR _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
