On Wed, Oct 15, 2008 at 11:15 PM, Andrew Beekhof <[EMAIL PROTECTED]> wrote: > > On Oct 15, 2008, at 11:36 PM, Serge Dubrouski wrote: > >>> + Implement a stack-independant version of pingd - no need to add ping >>> nodes to ha.cf >> >> Is it supposed to support the old way trough ha.cf? > > Yep > >> I still have problems there. > > Remind me?
With pingd in ha.cf: attrd[14279]: 2008/10/16_10:31:52 info: main: Starting mainloop... attrd[14279]: 2008/10/16_10:31:52 info: attrd_ha_callback: (null) message from (null) attrd[14279]: 2008/10/16_10:31:52 info: find_hash_entry: Ignoring message with no attribute name pingd[14273]: 2008/10/16_10:31:57 info: do_node_walk: Requesting the list of configured nodes pingd[14273]: 2008/10/16_10:31:59 info: send_update: 0 active ping nodes pingd[14273]: 2008/10/16_10:31:59 info: main: Starting pingd attrd[14279]: 2008/10/16_10:31:59 info: find_hash_entry: Creating hash entry for pingd attrd[14279]: 2008/10/16_10:32:00 info: attrd_trigger_update: Sending flush op to all hosts for: pingd attrd[14279]: 2008/10/16_10:32:00 info: attrd_ha_callback: (null) message from (null) attrd[14279]: 2008/10/16_10:32:00 info: find_hash_entry: Ignoring message with no attribute name Resources don't start. With pingd configured as a clone of RAs pingd attribute value doesn't get published anywhere and resources don't start again. > > > _______________________________________________ > Pacemaker mailing list > [email protected] > http://list.clusterlabs.org/mailman/listinfo/pacemaker > -- Serge Dubrouski. _______________________________________________ Pacemaker mailing list [email protected] http://list.clusterlabs.org/mailman/listinfo/pacemaker
