Hi! I am reading the corosync.conf man page. Seems that I have to increase token and token_retransmits_before_loss_const.
token This timeout specifies in milliseconds until a token loss is declared after not receiving a token. This is the time spent detecting a failure of a processor in the current configuration. Reforming a new configuration takes about 50 milliseconds in addition to this timeout. The default is 1000 milliseconds. token_retransmits_before_loss_constThis value identifies how many token retransmits should be attempted before forming a new configuration. If this value is set, retransmit and hold will be automatically calculated from retransmits_before_loss and token. The default is 4 retransmissions. I will try. Thanks! On Tue, Aug 13, 2013 at 12:35 AM, Arnold Krille <arn...@arnoldarts.de>wrote: > On Mon, 12 Aug 2013 19:27:33 +0200 Adrián López Tejedor > <adrian...@gmail.com> wrote: > > The problem is the network is out of my control. All the nodes are > > virtual machines over some VMWare ESX. > > We have two different networks, one for the service, and the other > > for the cluster. > > One idea is to create a second ring in the service network, but > > networks are virtualized, so maybe the problem persists. > > > > And of course, we don't have stonith. It is the same problem, I have > > no control over the VMWare hosts, and seems that they have to pay an > > extra to use the API needed by the stonith plugin. > > > > Meanwhile, I try to find > > > > Probably this two problems will be fixed in a couple of months, but > > meanwhile I have try to maintain the cluster up :) > > Sounds a bit as if the hardware-layer has bridging with stp featuring > some seconds of unavailability. > > You could try to increase the corosyncs parameters concerning > communication timeouts. > > Good luck, > > Arnold > > _______________________________________________ > Pacemaker mailing list: Pacemaker@oss.clusterlabs.org > http://oss.clusterlabs.org/mailman/listinfo/pacemaker > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org > >
_______________________________________________ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
