Scyld Nodes Freezing w/ SMP (fwd)

Nicholas Henke henken at seas.upenn.edu
Fri Nov 14 12:00:49 EST 2003


On Fri, 2003-11-14 at 11:31, Nicholas Henke wrote:
> On Fri, 2003-11-14 at 11:19, Jim Phillips wrote:
> > On Fri, 14 Nov 2003, Nicholas Henke wrote:
> > 
> > > On Fri, 2003-11-14 at 10:58, Jim Phillips wrote:
> > > > Hi,
> > > >
> > > > This is very similar to problems we're seeing on our dual Athlon MP 2600+
> > > > cluster with Gigabyte GA-7DPXDW+ motherboards, Intel PRO/1000 MT Server
> > > > network cards, and Clustermatic 3 (on Red Hat 8).  No solution, though.
> > >
> > > Does it work when using Lam/MPI ?
> > 
> > I don't know.  We run (and write) NAMD using direct TCP sockets.  Is Lam
> > bproc-compatible now?  Do you have some theory that the sockets code is
> > somehow responsible (other than simply stressing the machine)?
> 
> The 7.X tree of lam is bproc 'aware', and as far as why I would try it
> -- just another datapoint. It may help to isolate what exactly is
> causing the hangs.


BTW -- what version of bproc + kernel are you running ?

Nic
-- 
Nicholas Henke
Penguin Herder & Linux Cluster System Programmer
Liniac Project - Univ. of Pennsylvania

_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list