Community clusters moving to Slurm job scheduling system; training available throughout February and March

February 14, 2020

In the coming weeks, each community cluster supercomputer operated by ITaP Research Computing will transition to Slurm, a new batch job scheduling system.

Slurm is more scalable on large systems such as the community clusters than the current PBS-based system, and has become the state of the art for academic supercomputing centers. It is also open source, which will allow ITaP Research Computing to operate the community cluster program in the most cost-effective fashion.

The community cluster user guides have been updated with information about Slurm. The differences between Slurm and the current system, as well as how to convert scripts to Slurm, are covered in a Slurm Quick Reference Guide published by ITaP Research Computing.

ITaP Research Computing is offering hands-on training and help with Slurm throughout February and March from 2-3:30 p.m. on Fridays in the Envision Center, located off the hallway between the Purdue Memorial Union and Stewart Center. Interested individuals can RSVP now for the February 21 Slurm Transition Training, February 28 Slurm Transition Training and March 6 Slurm Transition Training.

In-person assistance is also available at Research Computing’s Coffee Hour Consultations, which now take place Monday through Thursday from 2-3 p.m. at various locations on campus.

Users who want to try out Slurm can already log into a testing environment. Access this testing cluster by ThinLinc at https://desktop.mack.rcac.purdue.edu or by SSH to mack.rcac.purdue.edu. This environment is small and intended for testing scripts with Slurm, and users should not do serious work on the testing cluster.

Email rcac-help@purdue.edu with any questions about the Slurm transition.

Originally posted: February 14, 2020  9:39am