Koushik, buch of fixes + make sure to update libqb (there is also bunch of fixes). Also 2.0, 2.1 and 2.2 are insecure (invalid initialization of crypto, there is CVE) so I would really recommend switching to supported 2.3.2 (plan is to support 2.3 series for quite a long time). Regards, Honza Koushik Sampath napsal(a): > Hi Jan, > > The corosync version: > Corosync Cluster Engine, version '2.1.0.27-232f' > > We have picked up a few bug fixes post the stable but my guess is that > there were more instabilities fixed. > > Is there any specific problem or is it a bunch of fixes that resolved this > problem? > > Thanks for the guidance! > Koushik > > > > > On 10 October 2013 00:52, Jan Friesse <jfriesse@xxxxxxxxxx> wrote: > >> Koushik, >> what version of corosync and libqb are you using? Can you please try >> latest one? There were various problems in that area in corosync < 2.3.1 >> and libqb < 0.15.0. >> >> Regards, >> Honza >> >> Koushik Sampath napsal(a): >>> What exactly do these log messages mean? Does it mean a cluster >> connection >>> failure? Can it be considered fatal with no expectation of a >>> reconnect? >>> >>> I have a functioning cluster which breaks after a while with these log >>> messages: >>> >>> Sep 18 19:26:20 notice [MAIN ] main.c:273 Completed service >>> synchronization, ready to provide service. >>> Sep 18 19:29:46 info [MAIN ] ipc_glue.c:490 Q empty, queued:0 sent:1. >>> Sep 18 19:29:49 error [MAIN ] ipc_glue.c:526 event_send retuned -146, >>> expected 13252! >>> Sep 18 19:29:49 error [MAIN ] ipc_glue.c:526 event_send retuned -134, >>> expected 232! >>> Sep 18 19:29:49 error [MAIN ] ipc_glue.c:526 event_send retuned -134, >>> expected 232! >>> >>> >>> Thanks >>> Koushik >>> >>> >>> >>> _______________________________________________ >>> discuss mailing list >>> discuss@xxxxxxxxxxxx >>> http://lists.corosync.org/mailman/listinfo/discuss >>> >> >> > _______________________________________________ discuss mailing list discuss@xxxxxxxxxxxx http://lists.corosync.org/mailman/listinfo/discuss
