[slurm-users] spreading jobs out across the cluster

[Stephen Berg, Code 7309]

I'm currently testing a new slurm setup before converting an existing pbs/torque grid over.  Right now I've got 8 nodes in one partition, 48 cores on each.  There's a second partition of older systems configured as 4 core nodes so the users can run some serial jobs.

--
Stephen Berg, IT Specialist, Ocean Sciences Division, Code 7309
Naval Research Laboratory
W:   (228) 688-5738
DSN: (312) 823-5738
C:   (228) 365-0162
Email: stephen.b...@nrlssc.navy.mil  <- (Preferred contact)
Flank Speed: stephen.p.berg....@us.navy.mil

During some testing I've noticed that jobs always seem to take the nodes in a top down fashion.  If I queue up a bunch of 3 node jobs they take nodes 1, 2 and 3 for one job, 4,5 and 6 for another. Nodes 7 and 8 never get used.  I'd like to have slurm spread the jobs out across the nodes in a round robin fashion or even randomly.  My config is really basic right now, I'm using defaults for most everything. Which settings could get the jobs spread out across the nodes in each partition a bit more fairly?

smime.p7s
Description: S/MIME Cryptographic Signature