Resource utilization exceeds limits


CentOS 7
PBSPro 14.1.0

I am seeing the following log entry in regards to some of the jobs run on my cluster:

ncpus 2.27 exceeded limit 1 (sum)

The jobs continue to run, but would it be possible to configure the scheduler to kill jobs that exceed their requested resources?

Thank you,



Hi SS,

If you had $enforce cpuaverage in your MoM’s config file - your job would have been aborted.
For various limits that you can apply on a job’s resource usage - you can go through “Placing Resource Limits on Jobs” in the Administrator Guide.



That’ll do the trick.

Thank you,