[Pacemaker] Combined colocation and order dependencies

Andrew Beekhof andrew at beekhof.net
Thu Dec 6 21:11:43 EST 2012


On Thu, Dec 6, 2012 at 10:27 PM, Lars Marowsky-Bree <lmb at suse.com> wrote:
> On 2012-12-06T20:11:41, Andrew Beekhof <andrew at beekhof.net> wrote:
>
>> > In my head it's always been that kind of (join|whatever) statement with order
>> > and/or colocation as attributes, that can be optionally turned off. LCMC
>> > presents it this way, but it's lot of pain, especially the resource sets are
>> > tricky.
>> So is that a vote for "too hard, do it in the XML" ?
>
> I'd vote for that, at least. Figuring out if two graphs (order and colo)
> match to the point where they can be merged is non-trivial to implement,
> and it is not something we'd want multiple frontends to have to do, I
> think.
>
> Resource sets actually go a long way in this direction already.
> (order with a set and sequential=true)

sequential=true has nothing to do with colocation, it just turns on
ordering within the set.
pretty sure its the default

>
> I actually wonder if that may not already solve the issue (wouldn't that
> be neat ;-), or if something is still missing.
>
>
> 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
>
>
> _______________________________________________
> 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://bugs.clusterlabs.org




More information about the Pacemaker mailing list