[Pacemaker] Static vs dynamic IPs

[Lars Marowsky-Bree]

On 2012-10-02T19:28:28, Andrew Beekhof <andrew at beekhof.net> wrote:

> > What are the implications of running Pacemaker with dynamic IP nodes? The IP
> > would change on restart and all configuration would be wrong. Can it be done
> > at all?
> Nope.  Not because of pacemaker, but the underlying messaging layer
> (heartbeat/corosync) wouldn't be able to find its peer when it comes
> back up.

As long as it's corosync with multicast (and the nodes still being able
to communicate on the mcastaddr), it ought to be possible there.

Mind, it isn't necessarily a great idea, but it should work. I'm not
sure if statically assigned nodeids are going to be more reliable or
not though.


Regards,
    Lars

-- 
Architect Storage/HA
SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 21284 (AG Nürnberg)
"Experience is the name everyone gives to their mistakes." -- Oscar Wilde