Forgot to CC the list.
---------- Forwarded message ---------- From: Dan Frincu <df.clus...@gmail.com> Date: Mon, Feb 6, 2012 at 10:16 PM Subject: Re: [Linux-HA] cluster hangs on maintenance-mode=false after updates on Ubuntu Lucid To: Walter Feddern <wfedd...@rogers.com> Hi, On Mon, Feb 6, 2012 at 10:09 PM, Walter Feddern <wfedd...@rogers.com> wrote: > On Mon, Feb 6, 2012 at 9:48 PM, Dan Frincu <df.cluster at gmail.com >> wrote: >> Hi, >> >> On Mon, Feb 6, 2012 at 9:27 PM, Walter Feddern <wfeddern at rogers.com> >> wrote: >>> We are running Ubuntu 10.04 (lucid). I am trying to upgrade from the base >>> version of corosync and pacemaker, to the most recent available via >>> https://launchpad.net/~ubuntu-ha-maintainers/+archive/ppa >>> >>> This is a test cluster, but I am trying to determine the best way to >>> update without interrupting the running resources. Here is the commands I am >>> using (on each node, with the exception of the first and last commands): >>> >>> crm configure property maintenance-mode=true >>> apt-get install python-software-properties >>> >>> add-apt-repository ppa:ubuntu-ha-maintainers\/ppa >> Second command. ^^ >>> apt-get update >> Third ^^ >>> /etc/init.d/corosync stop >> Fourth ^^ >>> apt-get install corosync pacemaker >> Fifth ^^ >>> crm configure property maintenance-mode=false >> This ^^ is the first command you need to run. >> >>Jumped ahead a bit, this needs to be set to true (e.g. you put the >>cluster in maintenance first, then upgrade corosync and pacemaker).. >> >> >> Afterwards you start corosync, check with ptest if the resources are >> fine, and then remove the maintenance-mode. >> >>e.g.: crm configure property maintenance-mode=false > > > > Not quite following what you are saying. at the start I run: > > crm configure property maintenance-mode=true > > which puts the cluster in maintenance mode while I do the updates. At this > point resources show as (unmanaged) > > after the update procedure completes (the update starts corosync again), and > all cluster nodes show online, and the resources show as started and in > (unmanaged) mode, I run: > > crm configure property maintenance-mode=false > > to try to take it out of maintenance mode and start monitoring the resources > again, but the command hangs and never completes. Yeah, you're probably hitting a known bug at this point. See http://www.mail-archive.com/pacemaker@oss.clusterlabs.org/msg11072.html > -- Dan Frincu CCNA, RHCE -- Dan Frincu CCNA, RHCE _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
