[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha-dev
Subject:    Re:  [Linux-ha-dev] RE: A bit of design specs of the diagnostic andrepair framework
From:       Lars Marowsky-Bree <lmb () suse ! de>
Date:       2000-10-08 17:40:01
[Download RAW message or body]

On 2000-10-05T08:10:48,
   David Brower <david.brower@oracle.com> said:

> Does anyone close to it think there is enough of a 
> reusable event system in FailSafe use that as a model?

FailSafe does have the concept of "process groups", which processes throughout
the cluster can join. (The so called group communication service).

This does provide reliable, ordered messaging.

This could conceivably be used to be used as a reusable event passing system,
if I understand your question correctly. It would of course be up to the
application on how to react.

The GCS _does_ have some performance issues at this point in time, as the
underlying algorithm was designed for HIPPI as a cluster interconnect (ie it
only uses point to point messaging instead of multicast, which would be quite
an improvement for mcast capable networks, patches are being accepted).

Sincerely,
    Lars Marowsky-Brée <lmb@suse.de>
    Development HA

-- 
Perfection is our goal, excellence will be tolerated. -- J. Yahl


_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.tummy.com
http://lists.tummy.com/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/

[prev in list] [next in list] [prev in thread] [next in thread]