Changes between Version 12 and Version 13 of Workshops/JobCheckpointing

Timestamp:

09/03/2026 11:35:20 AM (2 days ago)

Author:

Carl Baribault

Comment:

Fixed grammar, added long computation incentive

Workshops/JobCheckpointing

@@ -17 +17 @@
 == Pro's of job checkpointing ==
 
+* '''Checkpointed computations can run ''longer than cluster walltime limits''.''' See '''--qos=long''' in [wiki:cypress/about#SLURMresourcemanager SLURM (resource manager)].
+
 * '''Checkpointed jobs can get started sooner''' out of the job queue pending state with a reduced requested run time. (See "backfill scheduling" in [https://slurm.schedmd.com/sched_config.html|SLURM Scheduling Configuration Guide].
 * '''More checkpointed jobs can run simultaneously''' due to strict limits enforced by Cypress, LONI, and most other production clusters.
- * See see '''--qos=normal''' in [wiki:cypress/about#SLURMresourcemanager SLURM (resource manager)].
+ * See '''--qos=normal''' in [wiki:cypress/about#SLURMresourcemanager SLURM (resource manager)].
  * See also the command
    * '''sacctmgr show qos format=Name,!MaxWall,!MaxNodesPerUser | grep -E "normal|long"'''
 * '''Checkpointing mitigates the loss of resources and data due to either hardware failures or infinite looping jobs ''' - especially for otherwise long running parallel MPI jobs.
 * '''Checkpointed jobs can handle frequent job pre-emption''' - especially for certain cloud-based job queues with high availability.
-* '''Easier maintenance for HPC cluster managers''' - shorter maintenance periods waiting for nodes to become idle for servicing and reduced risk from less need for postponement of updates for security, hardware, and software.
+* '''Easier maintenance for HPC cluster managers''' - shorter maintenance periods waiting for nodes to become idle for servicing and reduced risk with quicker updates for security, hardware, and software.