On Fri, Jul 27, 2007 at 08:34:40PM +0200, Mark Hlawatschek wrote:
> On Friday 27 July 2007 15:45:19 Lon Hohberger wrote:
> > I think there's already a feature request open about it, by you?
> Yes, there are two feature requests open that address service dependencies
> 247980 - RFE: (strong and weak) service dependencies
> 247772 - RFE: One service following another
> >
> > Implementation-wise, in RHEL5, there's an event queue for service state
> > changes, and a function which broadcasts starts/stops
> > (broadcast_event()). The difference is that in RHEL5, you can multicast
> > to all rgmanagers w/ one call - whereas on RHEL4, you have to open a
> > connection to each and send the message individually.
> >
> > So, basically, it involves porting the event queue code and implementing
> > broadcast_event().
> OK, how much effort - in your opinion - would it be to port/implement these
> features ?
'follow' is harder to implement cleanly than strong vs. weak deps;
mostly because it's a two-step process. On RHEL5, it's easier due to
the event mechanism (you can trigger off 'service started' statechange
and move the conflicting service to another node).
The easiest path would be to port the event mechanism to RHEL4.
So, it's more likely to appear in RHEL5 first (less work).
The current train of thought is something like this:
* A follows B
* When A transitions to started, move B to another
node if they are on the same node.
> Is there a task scheduled to do these things ?
I can find out.
I assume you're willing to assist in testing, is that right?
--
Lon Hohberger - Software Engineer - Red Hat, Inc.
--
Linux-cluster mailing list
Linux-cluster@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/linux-cluster
