[Debian-ha-maintainers] Bug#553166: Bug#553166: closed by Guido Günther <agx at sigxcpu.org> (Re: Bug#553166: redhat-cluster: services are not relocated when a node fails)

[Guido Günther]

On Fri, Oct 30, 2009 at 09:13:50AM +0000, Martin Waite wrote:
> Hi Guido
> 
> On Thu, Oct 29, 2009 at 6:10 PM, Guido Günther <agx at sigxcpu.org> wrote:
> > Hi Martin,
> > On Thu, Oct 29, 2009 at 04:56:17PM +0000, Martin Waite wrote:
> >> Hi Guido,
> >>
> >> I have abandoned using RHCS on Debian Stable.   RHCS 2 in Fedora and
> >> Centos do not have the bug, and unfortunately in my production
> >> environment I will be tied to using RHCS 2.
> >>
> >> I do believe that the bug has gone away in version 3 of RHCS, but
> >> version 2 of RHCS as supplied by Lenny is virtually useless:  it
> >> cannot handle failover.
> >>
> >> As a newcomer to using RHCS, it took a few days to figure out that the
> >> problems I was having were not caused by my configuration errors (of
> >> which I had many) but were actually caused by a bug.
> >>
> >> Is there some way of either removing the package or providing some
> >> warning to potential users that it doesn't work ?   Marking the bug as
> >> closed seems inappropriate because it implies that the bug has been
> >> resolved - which is not true.    Would "wont fix" be better ?
> > The bug is fixed in RHCS 3.0.2 so closing it is apropriate. I wouldn't
> > object to remove RHCS 2 from Lenny though since I never got anything to
> > work with RHCS2 either (neither rgmanager nor gfs). I'm cc'ing the
> > maintainers of the RHCS2 package in Lenny. Is there anybody really using
> > RHCS 2 in Lenny in production? If not we should remove it.
> > Cheers,
> >  -- Guido
> >
> 
> On further thought, it is probably best to just leave things as they
> are.  From exchanges with the linux-cluster mailing list, some people
> do use the lenny package - but patch and rebuild the source package
> themselves.
Can you identify the patch? If it's small we can fix this up in a point
release. Help on this would really be appreciated since I'm not running
Cluster 2.
Cheeers,
 -- Guido