Skip to content

greenape/pbs_bullet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

80 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PBSBullet

A simple python module for submitting, and monitoring PBS jobs, and killing them if they run low on free memory. By default, the script will check on the job every 5 minutes.

If provided with a Pushbullet API token, it can also send notifications on job status.

Can be installed with pip -

pip install git+https://github.com/greenape/pbs_bullet.git

(Although you may need to use the --user switch, depending on your setup.)

Usage:

#Use pushbullet to notify on start, kill, and finish. Terminate if there is less
#than 10% free memory on any one node.
pbs-bullet --pushbullet-token <your api token> --low-mem 10 <jobnumber>

#Notify, but only on kill
pbs-bullet --pushbullet-token <your api token> --low-mem 10 --notify-on kill <jobnumber>

#Don't notify, and check every minute.
pbs-bullet --low-mem 10 --poll-interval 60 <jobnumber>

pbs-bullet can also take a job script as an argument with the --submit switch. If this option is used, the script expects a path to a job script instead of a job number. It will attempt to submit the script, and then monitor it.

#Submit a job script, notify using pushbullet, and kill the job if free memory on any nodes is below 10%
pbs-bullet --pushbullet-token <your api token> --low-mem 10 --submit /path/to/your/script.pbs

(You can also run the python module directly with python -m pbsbullet.pbs_bullet.)

Use this at your own risk: I have tested this only briefly, and only on Southampton's IRIDIS. I have no idea if it will work in other PBS environments, and I'd strongly suggest exercising extreme caution.

About

PBS job watch script for use with Pushbullet

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages