Skip to Main Content

The University of Tennessee

Newton header banner

Frequently Used Tools:

Job run time

Each queue has specific policies that enforce how long jobs are allowed to execute: short* queues allow up to 2 hours, medium* queues allow up to 24 hours, and long* queues have no runtime limit. When you submit a job, you are either implicitly or explicitly indicating how long the job is expected to run. The first way to to indicate the maximum runtime using the format

#$ -l s_cpu=12:00:00 
#$ -N test
uname -a

This job files is requesting up to 12 hours of CPU time (wallclock time) per CPU. The Grid Engine may internally weight this value in order to take into consideration the clock speed of the CPUs on which the job is executing.

The second way to indicate job runtime is to submit the job to a specific queue:

#$ -q medium*
#$ -N test
uname -a

This will request that the job execute in any queue matching the wildcard pattern "medium*". This will effectively enforce the default "medium" queue runtime limit of 24 hours. This is the same as using the "-l s_cpu=24:00:00" option.

The third way to indicate job runtime is to allow the system to assign the default. Any job that does not indicate a preferred queue or s_cpu value will be executed in the "short*" queues with a 2 hour execution limit. Once the runtime limit has elapsed, the Grid Engine will send an XCPU signal to the job followed by a KILL signal. You can choose to trap this signal if you wish to end your job gracefully, otherwise the job is forcibly killed.

Back to Using The Grid Engine