[torquedev] I think someone broke trunk...

Glen Beane glen.beane at gmail.com
Wed Feb 20 18:11:01 MST 2008


On Wed, Feb 20, 2008 at 7:58 PM, Garrick Staples <garrick at usc.edu> wrote:

> On Wed, Feb 20, 2008 at 04:36:16PM -0800, Garrick Staples alleged:
> > On Wed, Feb 20, 2008 at 04:26:54PM -0800, Garrick Staples alleged:
> > > On Wed, Feb 20, 2008 at 07:21:18PM -0500, Glen Beane alleged:
> > > > On Wed, Feb 20, 2008 at 7:14 PM, Garrick Staples <garrick at usc.edu>
> wrote:
> > > >
> > > > > On Wed, Feb 20, 2008 at 06:50:24PM -0500, Glen Beane alleged:
> > > > > > I'm thinking its revision 1949, but I haven't looked at it too
> closely
> > > > > >
> > > > > >
> > > > > > basically once I start pbs_mom all my client commands fail with
> a
> > > > > "Premature
> > > > > > end of message" error, so I can run qsub, qstat, etc as long as
> pbs_mom
> > > > > > isn't running yet...
> > > > > >
> > > > > >
> > > > > > is anyone else seeing this with trunk now?
> > > > >
> > > > > Seems fine to me.
> > > > >
> > > >
> > > >
> > > > disabling unix  sockets fixed things for me
> > > > I think there is some kind of regression there
> > >
> > > I haven't made any changes to the unix socket code for a few weeks.
> > >
> > > (though an unrelated unix socket change may be checked in soon)
> >
> > Ah, I think I see what is happening.
> >
> > pbs_mom is opening a unix domain socket and I didn't even realize it.
>  So
> > client connections are going to pbs_mom instead of pbs_server!
>
> Just checked in an ugly temporary fix.  Please see if things are working
> again.
>

seems to be working for me
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/torquedev/attachments/20080220/0ffb4565/attachment.html


More information about the torquedev mailing list