Hi, On Thu, Jun 19, 2008 at 02:55:57PM -0400, sth wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi folks, > > Please pardon me if this has been answered a million times already, but > I've checked the FAQs I could find, did my best at searching the list > archives with Google, and searched the docs that came with Heartbeat, > but couldn't find an answer, either way: > > Is there a way to manually override the initdead setting?
Looking at the source it seems like that there's no way to do that. Another possibility may be to enable autojoin and remove the node which is certainly down. Though I'm not sure if that would help. > I've tried to > force an hb_takeover on a node that's sitting in its initdead, and it > seems to go unheard (as in, the fact that I've run it isn't logged, as > it would normally be and the takeover doesn't happen). Is there a signal > that can be sent to heartbeat to say, "Do this now, please?" > > The initdead period isn't such a long wait, but I'm looking an an > application where keeping it at 120 sec is likely to be preferable, and > am thinking ahead to situations wherein my secondary node is the only > one of the cluster that can be booted after a power loss. For this > particular application, I'd like to be able to manually intervene and > avoid the added two-minute outage. Depending on your network you may reduce initdead. > I realize that the machine's ha.cf could be altered, in a pinch, but > thought that the ability for the sysadmin to exercise their discretion > and tell heartbeat, "Don't wait for your friend: she's not coming back," > might be in there, somewhere. > > Anyway, thanks for your time. Thanks, Dejan _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
