On 09/24/2010 01:43 PM, Lars Kellogg-Stedman wrote: >> Please install the debuginfo package for corosync from the repo, and then >> attach to the process with gdb: > > Steve, > > Thanks for your reply. I've attached the output to this email. > > For what it's worth: this seems to be happening every time I reboot > one of the cluster nodes. If after I boot the system I "killall -9 > corosync" and then "service corosync start", it works fine. > > -- Lars
Lars, pacemaker is waiting for something in nanosleep. Not sure what. The symptom you describe sounds like a inability for corosync to form a membership because of switch-default STP settings. One thing you could try is to enable fast STP in your switch config or use broadcast mode (which avoids STP startup times). Try running the following on the node after a lockup: killall -SEGV corosync corosync-fplay attach output Regards -steve _______________________________________________ Openais mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/openais
