[torqueusers] Updated: killbaduser, a tool to clean up rogue user processes

Ole Holm Nielsen Ole.H.Nielsen at fysik.dtu.dk
Wed Nov 1 14:10:49 MST 2006


Regarding getting job information on Torque nodes, I asked:
> Alternatively, it might be possible to extract the job-ID and user
> information from the *.JB files (it's in there, I've looked at an octal 
> dump).

Åke Sandgren ake.sandgren at hpc2n.umu.se wrote:
> printjob /var/spool/torque/mom_priv/jobs/7471.audhum.JB
and Garrick Staples wrote:
> Use 'printjob',
>   printjob /var/spool/torque/mom_priv/jobs/7471.audhum.JB | grep euser

Yes, the printjob command does the trick !  I've updated the killbaduser script
to version 1.4 and placed it for download from ftp://ftp.fysik.dtu.dk/pub/PBS/
Now the pbs_server isn't inquired any more, so killbaduser ought to
be very efficient !  The description is:

# On a Torque/PBS compute node, list and kill any user processes not belonging 
to batch jobs.
#
# Usage: killbaduser [-k] [-s] [-v]
#    -k will execute the kill command
#    -s will sleep a random number of seconds
#    -v verbose output for debugging
# Author: Ole Holm Nielsen, Department of Physics, Technical University of Denmark
# Version: 1.4


Thanks,
Ole


More information about the torqueusers mailing list