[Openais] call for roadmap features for future releases

[Jan Friesse]

Steven Dake wrote:
> I invite all of our contributors to help define the X.Y roadmap of both
> corosync and openais.  Please submit your ideas on this list.  Some
> examples of suggested ideas have been things like converting to libtool.
> Also new service engine ideas are highly welcome.  Keep ideas within a 1
> month - 3 year timeframe.
> 
> I intend to publish the roadmap with the release of Corosync and OpenAIS
> 1.0.0.  Please submit your ideas by June 26, 2009 (friday).
> 
> Regards
> -steve
> 
> _______________________________________________
> Openais mailing list
> Openais at lists.linux-foundation.org
> https://lists.linux-foundation.org/mailman/listinfo/openais

It should be nice to have some kind of availability manager (let's call
it SAM (simple availability manager)). It should be something to prevent
 application failure (or better, to be able restart registered service).

Please note, that this shouldn't be alternative to rgmanager/piecemaker
(so working for application, which are not developed to be monitored)
but more for AMF (application, which calls some functions and know about
it's monitoring). Problem of AMF is, that is super overcomplicated and
not finished. We should take API similar to component health monitoring
and implement only this part.

Engine should be only single node and detect failure of application. If
application will fall (check should be similar like watchdog), it is
restarted. So no services migration, ...

I think, this is missing part of CPG, because most of CPG applications
works like replicated finite automata. If cpg application fails, this
finite automata can lock up. Simple availability manager can solve it by
restarting process.

Regards,
  Honza