Hi, On Wed, Nov 04, 2009 at 01:30:21PM +0000, S. A. Woltering wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Dejan Muhamedagic wrote: > >> "button -S" is not supported. Only "button" and "power". The > >> latter is, I think, more reliable, the former easier on the > >> hardware. And riloe should issue a warning if it doesn't > >> recognize the method. > > Curiously, "-S" works. This option seems to power off and KEEP it off. > Which is, of course exactly the behaviour I'm after. > > I've tried it without the S-option and the node restarts.
If external/riloe doesn't recognize the method, then it defaults to "power" (though silently), so with "button -S" you actually get "power". > >> You have to cleanup resources after manual intervention. > > Will do. But I had a curious issue whereby "cat /proc/drbd" was showing > all the right things, but "crm_mon" was still claiming the secondary > node was disconnected. Unfortunately, those are two different things. The split brain recovery is done separately in both drbd and the cluster stack (openais). > After cleaning up the resources it still took a > long time to sort itself out. That shouldn't have happened. You can file a bugzilla with hb_report for further investigation. > >> This is two issues: a) Why the drbd resource can't start after > >> split-brain and b) Why does it take such a long time for the > >> cluster to recover after resource recovery and cleanup. > > Well, since my last lot of cleanup, it is now behaving itself and my > tests are getting increasingly aggressive. It's not reproducible? Happened only once? Thanks, Dejan > Not sure why it took a time to settle down, but if an hb_report would > help I'll ship one. > Thanks, > A > - -- > Ashley Woltering, Systems Analyst > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v2.0.4-svn0 (GNU/Linux) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iD8DBQFK8YHth854NVK99FMRAmB0AJ9uDHK0Jjq4Ofr/uydd6YoJV+j6iQCffkKW > oFxDFp067gxLUhb4p6NJklc= > =sg81 > -----END PGP SIGNATURE----- > _______________________________________________ > 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
