Hi, On Tue, Feb 12, 2008 at 12:16:13PM -0500, [EMAIL PROTECTED] wrote: > I have been using the mysql OCF on heartbeat-2.1.2-3 for some time on centOS > 5.1 with no problems > > Feb 12 12:11:42 utnode1 haclient: on_event:evt:cib_changed > Feb 12 12:11:44 utnode1 mysql[12728]: [12807]: INFO: MySQL started > Feb 12 12:11:44 utnode1 crmd: [12025]: info: process_lrm_event: LRM operation > mysql_repl_start_0 (call=27, rc=0) complete > Feb 12 12:11:44 utnode1 crmd: [12025]: info: build_operation_update: Digest > for 0:0;10:37:50df748e-1a0e-45b3-ad16-0dd37fbf9450 (mysql_repl_start_0) was > 6caa97789ad6bbd931da5b3e8677b611 > Feb 12 12:11:44 utnode1 crmd: [12025]: info: log_data_element: > build_operation_update: digest:source <parameters target_role="started" > datadir="/replicated/mysql/data"/> > Feb 12 12:11:46 utnode1 haclient: on_event:evt:cib_changed > > When attempting the start up after upgrading all other resources start with > no problems but when it gets to MySQL I get > > Feb 12 12:00:50 utnode1 lrmd: [11257]: WARN: operation start[14] on > ocf::mysql::mysql_repl for client 11260, its parameters: > target_role=[started] CRM_meta_timeout=[20000] crm_feature_set=[1.0.9] > datadir=[/replicated/mysql/data] : pid [11555] timed out > Feb 12 12:00:50 utnode1 crmd: [11260]: ERROR: process_lrm_event: LRM > operation mysql_repl_start_0 (14) Timed Out (timeout=20000ms)
It's hard to say why the mysql script timed out. Probably network related. You could try to run the mysql RA by hand: # export OCF_ROOT=/usr/lib/ocf # OCF_RESKEY_datadir="/replicated/mysql/data" sh -x /usr/lib/ocf/resource.d/heartbeat/mysql and see where it blocks. > Feb 12 12:00:52 utnode1 crmd: [11260]: info: do_lrm_rsc_op: Performing > op=mysql_repl_stop_0 key=1:25:50df748e-1a0e-45b3-ad16-0dd37fbf9450) > Feb 12 12:00:52 utnode1 lrmd: [11257]: info: rsc:mysql_repl: stop > > Reinstalling the previous heartbeat version restored > functionality with no changes to the cib.xml made or to > anything else besides the heartbeat RPMs. That means that there's a regression in the mysql RA. Just looked through the changelog and there's a change in how uid are resolved: instead of grep /etc/passwd, getent is used. Where are the security databases (passwd/group)? How's your nsswitch.conf setup? Thanks, Dejan > > Any help appreciated, > > Amy > > _______________________________________________ > 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
