[torqueusers] Job output files not being copied to head

Lippert, Kenneth B. Kenneth.Lippert at alcoa.com
Wed Apr 4 06:44:50 MDT 2007


I have found this almost always means the compute nodes cannot ssh or
rlogin to the host.  Make sure all nodes and the host can login to each
other without passwords.  Even if using NFS, the compute node still
wants to login to return the log files.

-k




-----Original Message-----
From: torqueusers-bounces at supercluster.org
[mailto:torqueusers-bounces at supercluster.org] On Behalf Of
aohara at haverford.edu
Sent: Wednesday, April 04, 2007 2:15 AM
To: torqueusers at supercluster.org
Subject: [torqueusers] Job output files not being copied to head

Hi,

I'm using TORQUE vers. 2.1.8 with Maui vers. 3.2.6p19.  The cluster is 6
node 48 processor (4 dual core amd64 processors per node) running Ubuntu
Dapper Drake with a 2 processor (dual core) head node, also running
Ubuntu.

The problem is that after running jobs, all of the output files (e.g.
44.babbage.OU and 44.babbage.ER) remain in the undelivered directory on
the compute nodes rather than being copied into the submission director
on
the head node.  I have tried with the default configured scp and with
the
added line from the manual for using nfs and neither work.  Anybody have
some hints?  I'd prefer to be able to use nfs in the end.

Thanks in advance,
Andy O'Hara
_______________________________________________
torqueusers mailing list
torqueusers at supercluster.org
http://www.supercluster.org/mailman/listinfo/torqueusers


More information about the torqueusers mailing list