Intro to SLURM Workshops

Published May 30, 2013 This content is archived.

CCR Staff will host introductory sessions on the SLURM Workload Manager throughout the summer.

Print
“The Simple Linux Utility for Resource Management (SLURM) is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small Linux clusters. ”
SchedMD Website

All sessions are from 3-4pm in Clemens 120 on UB's North Campus:

Wednesday, July 10, 2013
Wednesday, September 4, 2013

Over the last year CCR staff has been configuring and testing a replacement to our batch job and scheduling software, PBS/Torque.  We are pleased to be in final preparations for the transition to the SLURM workload manager.  SLURM has been in use for some time at DoE HPC facilities, and is becoming increasingly popular at academic sites.   We've had several large faculty groups testing SLURM at CCR over the last few months and we're confident you will be happy with the switch. 

Early in June, half of the u2 cluster was brought into the new SLURM configuration.  On Tuesday, July 9th the remaining half of the U2 cluster was migrated into SLURM. 
 
CCR Staff will be holding  informational sessions for all CCR users so you can get an idea of what SLURM is and how the new system is setup.  The presentation will cover topics such as:

  • What is SLURM and why we chose it
  • What has changed
  • What resources are now available in the CCR cluster
  • Examples of how to run jobs in SLURM
  • Job Monitoring
  • How to get more information and help
  • Tools to make the migration easier

The presentation will be followed up by a question and answer period.  The sessions are free and open to all CCR users.  No pre-registration is required.  If you have any questions, please contact CCR Help.