[Pacemaker] Stopping corosync on the slave in a 2 node cluster causes all the resources on the master to stop. Please Help.

[Tom Pride]

Hi Florian,

I've just had a chance to try your suggestions.  The interleave option
didn't appear to make a difference, but getting rid of the order
ActiveMQ_Data_after_ping_gateway_clone constraint has done the job
perfectly.  I don't think I can see any real reason why I needed
that constraint.  I guess I put it in there to be more explicit about
ordering.

The reason I didn't put cluster_IP and drbd_fs in a group was to make it
easier for new broker resources to be added to the cluster config.  The
brokers are installed via rpm and the rpm executes some cibadmin commands to
automate the addition of the resource to the crm config which includes
adding an order constraint that look like this:

order activemq-emp_after_drbd_fs inf: drbd_fs:start activemq-emp:start

I can't remember the exact specifics now, but I think it made it difficult
to automate this if we used a group.

Cheers,
Tom


On Wed, Oct 12, 2011 at 6:23 PM, Florian Haas <florian at hastexo.com> wrote:

> Tom,
>
> acting on a hunch here: do things improve if you add "meta
> interleave=true" to your ping_gateway_clone, or if you remove the
> ActiveMQ_Data_after_ping_gateway_clone order constraint?
>
> If that doesn't get you anywhere, could you produce a CIB dump of the
> cluster when your one node is shut down, i.e. when it's in the error
> state you're describing. Please don't paste inline this time, instead
> please create the CIB dump and upload it to
> pastebin/pastie/fpaste/whathaveyou.
>
> Also, as an aside, is there a specific reason why you didn't put
> cluster_IP and drbd_fs in a group? That would make the configuration a
> little more compact.
>
> Cheers,
> Florian
>
> --
> Need help with Pacemaker?
> http://www.hastexo.com/now
>
> _______________________________________________
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20111014/d19432a3/attachment-0003.html>