[Linux-cluster] rgmanager fail to start a 'clone' service, is this a bug or expected behavior?

[Lon Hohberger]

On Fri, 2007-01-26 at 08:07 -0800, Roger PeXa Escobio wrote:
> Hi
> 
> I am just testing an active-active apache services
> with Redhat Cluster suit but rgmanager fail to start
> the httpd services in the second server althougt it
> claim it succefully started it ...
> 
> I filled a bug :
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=224462

> We are running RHEL4.4ES with gfscs from centos-4.4
> 
> so, is this a bug or an expected behavior?

Mostly, it's expected behavior; see lengthy explanation and example
outputs of rg_test in the bug.  However:

(a) rgmanager should log resource collisions, and
(b) it might be worth investigating having rgmanager figure out which
resources will never collide and allow reuse of certain "unique"
parameters (e.g. mountpoint, in your particular case).

-- Lon