-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 16/05/14 00:03, Schmidtmann, Carl wrote:
> How do you specify the node and partition in the slurm.conf file > to allow for each gpu to have one cpu core reserved? I remember > the answer is to have two partitions where one allows all but one > cpu core to be used as compute resources and then another partition > that has the gpu and all cores available. I have seen this given > out as an example but I can't find that. Maybe it could be included > on the gres page of the documentation. It's mentioned in the Slurm 2.6 slides from the Slurm User Group last year: http://slurm.schedmd.com/SUG13/version_26.pdf # GPU Scheduling Example # # For cluster with 1 GPU and 4 CPUs per node # # Configure 2 queues: “cpu” and “gpu” # # Configure queue “cpu” with MaxCPUsPerNode=3 # # Queue “gpu” use restricted to jobs using GPUs (enforce using # job_submit plugin) Is that what you were after? cheers, 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 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iEYEARECAAYFAlOGvCgACgkQO2KABBYQAh/jXwCfSLZ4dMk60YOuZeSLrsG7Kwuv hCUAmgMEvY7PVzZk2iAcxWFNAJ9gXKcQ =VqDE -----END PGP SIGNATURE-----
