technical:slurm:scheduler-params

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
Next revisionBoth sides next revision
technical:slurm:scheduler-params [2020-02-20 14:52] – freytechnical:slurm:scheduler-params [2020-02-26 15:20] – [Implementation] frey
Line 27: Line 27:
 |fair-share| 4000|see ''sshare''| |fair-share| 4000|see ''sshare''|
 |partition id| 2000|1.0 for all partitions| |partition id| 2000|1.0 for all partitions|
-|job size| 1|all cores in cluster=1.0|+|job size| 1|all resources in partition=1.0|
  
 Next to priority access, wait time is the largest factor:  the longer a job waits to execute, the higher its priority to be scheduled.  This seems appropriate, but it competes against the fair-share factor, which prioritizes jobs for underserved users. Next to priority access, wait time is the largest factor:  the longer a job waits to execute, the higher its priority to be scheduled.  This seems appropriate, but it competes against the fair-share factor, which prioritizes jobs for underserved users.
Line 76: Line 76:
  
 ===== Implementation ===== ===== Implementation =====
 +
 +The Priority weight factors will be adjusted as follows:
  
 ^Configuration Key^Old Value^New Value^ ^Configuration Key^Old Value^New Value^
-| ''PriorityWeightAge''| 8000| 2457| +|''PriorityWeightAge''| 8000| 2457| 
-| ''PriorityWeightFairshare''| 4000| 1073737728| +|''PriorityWeightFairshare''| 4000| 1073737728| 
-| ''PriorityWeightJobSize''| 1| 0| +|''PriorityWeightJobSize''| 1| 0| 
-| ''PriorityWeightPartition''| 2000| 0| +|''PriorityWeightPartition''| 2000| 0| 
-| ''PriorityWeightQOS''| 20000| 13835058055282163712| +|''PriorityWeightQOS''| 20000| 3221225472| 
-| ''PriorityWeightTRES''| unset| ''cpu=819,mem=245,gpu=245,node=327''|+|''PriorityWeightTRES''| unset| ''cpu=819,mem=245,GRES/gpu=245,node=327''| 
 +|''PriorityFlags''| ''FAIR_TREE,SMALL_RELATIVE_TO_TIME''| ''FAIR_TREE''| 
 + 
 +Priorities will make use of the full 32-bits available, with the fairshare factor dominating and having the most precision assigned to it. 
 + 
 +Since the ''PriorityWeightJobSize'' will not be used, the more complex "small-relative-to-time" algorithm will be disabled. 
 + 
 +The modifications to ''slurm.conf'' must be pushed to all systems.  The ''scontrol reconfigure'' command should be all that is required to activate the altered priority caclulation scheme.
  
 ===== Impact ===== ===== Impact =====
Line 93: Line 102:
 ^Date ^Time ^Goal/Description ^ ^Date ^Time ^Goal/Description ^
 |2019-10-24| |Authoring of this document| |2019-10-24| |Authoring of this document|
 +|2020-03-09|09:00|Implementation|