[Pacemaker] stop problem and crm node delete nodename is bug?

[Dejan Muhamedagic]

Hi,

On Thu, Sep 30, 2010 at 01:38:27PM +0200, Andrew Beekhof wrote:
> On Wed, Sep 29, 2010 at 9:46 AM, jiaju liu <liujiaju86 at yahoo.com.cn> wrote:
> 
> > Date: Tue, 28 Sep 2010 12:27:47 +0200
> > From: Andrew Beekhof <andrew at beekhof.net<http://cn.mc157.mail.yahoo.com/mc/compose?to=andrew@beekhof.net>
> > >
> > To: The Pacemaker cluster resource manager
> >     <pacemaker at oss.clusterlabs.org<http://cn.mc157.mail.yahoo.com/mc/compose?to=pacemaker@oss.clusterlabs.org>
> > >
> > Subject: Re: [Pacemaker] pacemaker stop problem
> > Message-ID:
> >     <AANLkTikoqp6bpEcxChxRVqSAshhBW=1KsV7SHjXBSMBG at mail.gmail.com<http://cn.mc157.mail.yahoo.com/mc/compose?to=1KsV7SHjXBSMBG@mail.gmail.com>
> > >
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > On Tue, Sep 28, 2010 at 10:00 AM, jiaju liu <liujiaju86 at yahoo.com.cn<http://cn.mc157.mail.yahoo.com/mc/compose?to=liujiaju86@yahoo.com.cn>>
> > wrote:
> >
> > > hi guys
> > > I use  command service openais force-stop to stop openais, It ofen waste
> > a
> > > long time to stop or maybe run this command and no end. sometimes I
> > > use command service openais force-stop twice it will be ok, or I have to
> > > kill pocess. who has a better way to stop service.
> > >
> > >
> > More than likely openais is waiting for pacemaker, and pacemaker is waiting
> > for one of your cluster services to stop.
> > Figure out why thats taking so long and you'll solve the issue.
> >
> > Are there any command to turn off pacemaker???
> >
> 
> see the bit above where I said its not pacemaker's fault?
> 
> but anyway, try: crm node standby
> that will stop all the services and let pacemaker exit quickly
> 
> 
> 
> > and I find some problem about crm node delete nodename.
> > If your delete node is DC the cluster will not vote new DC, so DC is NONE.
> >
> 
> You're deleting the current DC?
> Uh, dont do that.
> In fact, don't delete any nodes that are still part of the cluster.
> The docs are pretty clear on that one.

The shell won't let you remove a node which is online. Perhaps
this release is a bit older and don't include the latest code.
IIRC, it was done for 1.0.8.

Thanks,

Dejan

> > Besides these ,although the cluster could not see this node ,the node could
> > use crm_mon to see cluster. and If filesystem resource is running on this
> > node, after execute command crm node delete nodename, the filesystem will
> > not umount, so this resource could not migrate to other node in cluster I
> > think this is a bug.
> >
> >
> > _______________________________________________
> > Pacemaker mailing list: Pacemaker at 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
> >
> >

> _______________________________________________
> Pacemaker mailing list: Pacemaker at 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker