Hi, On Tue, Nov 10, 2009 at 06:55:44PM +0500, Muhammad Sharfuddin wrote: > OS: SUSE Enterprise 10 SP 3 i586 (with online updates) > > there are only two ocf resources (Filesystem and IP address) in my test > cluster. > > nodes are > virtual-1 192.168.0.242 > virtual-2 192.168.0.243 > > ping node is 192.168.0.6. > > in my test cluster, resources never migrates to the other node, even > when the node(running resources) cant ping to the ping node.
There are no constraints in your CIB. You need to define location constraints based on the ping attribute. I suppose that there's some documentation on how to do that. Also, put some constraints to express resource/node preferences, otherwise your resource may move around when you don't want that. Thanks, Dejan > e.g if the resources are running on virtual-2, and using firewall on > pingnode(192.168.0.6), I block every thing/communication from the node > running resources i.e virtual-2(192.168.0.243), to the ping > node(192.168.0.6)... but resources never migrates to other > nodes(virtual-1 192.168.0.242), and vice-verca. > > resources were running on machine virtual-2, and I blocked communication > of virtual-2 with pingnode(192.168.0.6) > > # grep pingd /var/log/messages > Nov 10 17:20:48 virtual-2 pingd: [6198]: notice: pingd_lstatus_callback: > Status update: Ping node 192.168.0.6 now has status [dead] > Nov 10 17:20:48 virtual-2 pingd: [6198]: notice: pingd_nstatus_callback: > Status update: Ping node 192.168.0.6 now has status [dead] > Nov 10 17:20:48 virtual-2 pingd: [6198]: info: send_update: 0 active > ping nodes > > still all the resources are running/available to virtual-2 > > now I enable communication with ping node. > # grep pingd /var/log/messages > Nov 10 18:46:38 virtual-2 pingd: [6198]: notice: pingd_lstatus_callback: > Status update: Ping node 192.168.0.6 now has status [up] > Nov 10 18:46:38 virtual-2 pingd: [6198]: notice: pingd_nstatus_callback: > Status update: Ping node 192.168.0.6 now has status [up] > Nov 10 18:46:38 virtual-2 heartbeat: [6090]: info: Status update for > node 192.168.0.6: status ping > Nov 10 18:46:38 virtual-2 pingd: [6198]: info: send_update: 1 active > ping nodes > Nov 10 18:46:38 virtual-2 pingd: [6198]: notice: pingd_nstatus_callback: > Status update: Ping node 192.168.0.6 now has status [ping] > Nov 10 18:46:38 virtual-2 pingd: [6198]: info: send_update: 1 active > ping nodes > Nov 10 18:46:43 virtual-2 attrd: [6105]: info: attrd_trigger_update: > Sending flush op to all hosts for: pingd > > > please help me, so that if a node cant ping to ping node(s), then > resources automatically migrates to other node. > > I use hb_gui to configure the cluster. > > Thanks and Regards > > autojoin none > > crm true > > ucast eth0 192.168.0.242 > ucast eth0 192.168.0.243 > #bcast eth0 > > use_logd yes > > > ping 192.168.0.6 > deadping 20 > > deadtime 30 > warntime 10 > initdead 120 > > > node virtual-2 > node virtual-1 > > _______________________________________________ > 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
