[torqueusers] Submitting Jobs on alpha-linux cluster

[David LeBard]

We would ideally like to keep our old hardware, but I do agree with you
that the opterons are a nice replacement for the number-crunching
alphas.  In fact, we are no longer buying new alphas, but instead we
created an all new homogeneous cluster of opterons (Incidentally, the
OpenPBS/Torque on this cluster has no problems at all, though it has the
same type of MOM/Server configuration as the alphas.).

Back to the original question: What does this message from the queue
mean, and how do I get my jobs "unstuck" from this queue?

Message from qstat -f: "Not Running - PBS Error:  MSG=send failed,
STARTING"

Thanks,
David

On Mon, 2005-02-21 at 15:42, Nico Kadel-Garcia wrote:
> ----- Original Message ----- 
> From: "David LeBard" <david.lebard at asu.edu>
> To: <torqueusers at supercluster.org>
> Sent: Monday, February 21, 2005 4:13 PM
> Subject: Re: [torqueusers] Submitting Jobs on alpha-linux cluster
> 
> 
> > Update to what?  I thought 7.2 was the latest redhat linux for alpha
> > processors.
> 
> Update to Opteron CPU's so you can use an x86_64 distribution? Alphas are 
> really end-of-life for a lot of operating systems. There is something afoot 
> to port Fedora Core to it, mentioned at 
> http://www.redhat.com/archives/fedora-devel-list/2003-October/msg00159.html, 
> but I don't recommend Fedora Core for clusters because it tends to be a bit 
> too cutting edge. I also hope that you're keeping an eye on 
> www.fedoralegacy.org to get any available security updates, at least by 
> being able to recompile SRPM's as needed for security updates.
> 
> But I recognize that if you're stuck with the hardware, it may be tough to 
> upgrade. 
>