On 04/06/16 01:20, Jason Bacon wrote:

> We're having an issue with CPU binding when two jobs land on the same node.

Use Slurm's cgroup support to ensure that they are partitioned onto
separate cores and so cannot clobber each other.

http://slurm.schedmd.com/cgroups.html

It's what we've been doing here since we started using Slurm, and we
used Torque's cpuset functionality before that for the same reason.

All the best,
Chris
-- 
 Christopher Samuel        Senior Systems Administrator
 VLSCI - Victorian Life Sciences Computation Initiative
 Email: [email protected] Phone: +61 (0)3 903 55545
 http://www.vlsci.org.au/      http://twitter.com/vlsci

Reply via email to