[torqueusers] Upgrade torque and continue

Richard Walsh rbw at ahpcrc.org
Wed Dec 14 15:53:48 MST 2005

group hpc wrote:

> Dear all,
> Does anyone know how to upgrade torque server or mom from old version 
> without removing the current running jobs data, so the jobs can 
> continue run after performance upgrade.
> Also, how to reset the job id numbering?
> Thanks, 
> Josh


Complete the build. You probably need to do this in separate directories
for the head and compute nodes to get specific configurations.  While the
current set of jobs are running 'make install' the new version.  This 
will drop
the new binaries into place without distributing those running.  Now retart
(stop and start in sequence) the moms using the services script in 
The mom should be started with the -p option in the script to re-acquire
running jobs using polling (-p for polling).  Now stop and start the server.
The server should be stopped in the script with: qterm -t quick to leave 
jobs running. This should do it. 

Verify that all is well with pbsnodes -a | less which should show the 
nodes in the correct state (free or job-exclusive).  The communication 
of state
between the server and the moms may take a few moments.


>torqueusers mailing list
>torqueusers at supercluster.org


Richard B. Walsh

Project Manager
Network Computing Services, Inc.
Army High Performance Computing Research Center (AHPCRC)
rbw at ahpcrc.org  |  612.337.3467

This message (including any attachments) may contain proprietary or
privileged information, the use and disclosure of which is legally
restricted.  If you have received this message in error please notify
the sender by reply message, do not otherwise distribute it, and delete
this message, with all of its contents, from your files.

More information about the torqueusers mailing list