Where does the heartbeat 2.0.7 binaries come from? What packaging? What distribution?
It's starting to sound (to me at least ;) like some broken heartbeat package. Yan PS - it is not normal for the CIB to be rewritten every few seconds. It should be rewritten when something in the cluster state changes (resource or node status) - which should not happen for no reason. Anders Brownworth wrote: > So to add some color to this problem, I'm seeing these errors on box02: > > ERROR: Exiting pengine process 4689 dumped core > ERROR: crmdManagedChildDied:subsystems.c The pengine subsystem > terminated unexpectedly > ERROR: do_log:misc.c [[FSA]] Input I_ERROR from crmdManagedChildDied() > received in state (S_TRANSITION_ENGINE) > ERROR: do_recover:control.c Action A_RECOVER (0000000001000000) not > supported > ERROR: do_log:misc.c [[FSA]] Input I_STOP from do_recover() received > in state (S_RECOVERY) > > This is a 2 box setup migrating an IP and service (OpenSER) to run on > one of > the nodes at any given time. Box01 doesn't see any errors and just leaving > the nodes alone to work it out doesn't work. I am unclear on the > cib.xmlbecause I haven't seen it work any other way. Is it normal to > have that file > rewritten every 5 seconds or so? Seems like that's abnormal to me but can't > get that confirmed in the docs. > > I'm happy to grab any other information if it might help. Thanks. > > -Anders > > On 6/12/07, Anders Brownworth <[EMAIL PROTECTED]> wrote: >> >> Andrew, >> >> Sorry about the formatting. Attached are the /var/log/messages output of >> box01 and box02 as well as the cib.xml. Also, this is heartbeat version >> 2.0.7. >> >> As you can see in the messages, this isn't a problem that gets worked >> out. >> Both machines keep writing and re-writing their cib.xml files all day >> long. I gather this isn't normal. >> >> Thanks for any light you can shed. >> >> -Anders >> >> On 6/12/07, Andrew Beekhof < [EMAIL PROTECTED]> wrote: >> > >> > On 6/12/07, Anders Brownworth < [EMAIL PROTECTED]> wrote: >> > > Hi, >> > > >> > > I have a heartbeat v2 setup that I am trying to use to migrate >> OpenSER >> > and >> > > an IP address back and forth between 2 boxes. (box01 and box02) I >> > wrote an >> > > OCF for OpenSER and am using the heartbeat provided IPaddr to manage >> > the IP >> > > address. My OCF checks out with the ocf-tester script and my >> > > cib.xmlverifies with crm_verify -x /var/lib/heartbeat/crm/cib.xml. >> > > >> > > When I start both nodes with exactly the same configuration, they >> > fight >> > > about what the state of things and the first ERROR I get is: >> > > >> > > crmd: [3507]: ERROR: do_exit:control.c Could not recover from >> internal >> > error >> > > >> > > in the /var/log/messages of box02 and neither service is started. (IP >> > nor >> > > OpenSER) Both boxes give off a pile of info and warning messages that >> > don't >> > > seem to point me in a worthwhile direction. >> > > >> > > Both boxes seem to get to "info: main:attrd.c Starting mainloop..." >> > without >> > > any issues. But once they try to decide on who has what, they start >> > fighting >> > > and constantly rewriting their cib.xml files over and over unendedly. >> > (Is >> > > this normal?) crmd starts complaning about pengine dying on a signal >> > 14 and >> > > everyting pretty much goes to hell from there. >> > >> > it _should_ be able to recover. >> > >> > can you send your logs as attachments? (trying to read logs wrapped at >> > 80 chars is hell) >> > >> > you also didnt mention what version you're running >> > _______________________________________________ >> > Linux-HA mailing list >> > [email protected] >> > http://lists.linux-ha.org/mailman/listinfo/linux-ha >> > See also: http://linux-ha.org/ReportingProblems >> > >> >> >> >> -- >> -Anders >> ----------------------------------------------------------- >> Anders Brownworth >> [EMAIL PROTECTED] >> >> > > _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
