[torqueusers] cluster underused - single cpu jobs hold up parallel

Christopher Samuel samuel at unimelb.edu.au
Thu Mar 31 22:44:14 MDT 2011


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 01/04/11 02:59, Lydia Heck wrote:

> The parallel jobs are not schedulled. The scheduller
> is maui. Any idea what I am missing here?

This is what we are using at VLSCI to prefer wide jobs
and to prioritise on XFACTOR not the queue time (thus
preferring shorter jobs as their xfactor will increase
faster than longer jobs, helping to backfill the system).

This started off in our Maui config and is now in our
Moab configuration.

######################################################################
# Queuing policies
# - Use XFACTOR rather than queue time
# - Increase a jobs priority by 1 each time it's bypassed
# - Give more weight to larger jobs (based on number of cores requested)
######################################################################

QUEUETIMEWEIGHT 0

SERVICEWEIGHT 10
XFACTORWEIGHT 100
BYPASSWEIGHT 1

# Allow users to modify their priority downwards.
USERPRIOWEIGHT  1

RESWEIGHT 1
PROCWEIGHT 10


- -- 
    Christopher Samuel - Senior Systems Administrator
 VLSCI - Victorian Life Sciences Computation Initiative
 Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545
         http://www.vlsci.unimelb.edu.au/

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEUEARECAAYFAk2VWB4ACgkQO2KABBYQAh/1dwCfchg647lJzTS047y4/qpKdXOZ
S+sAmN6b1l4CFnYB/UKaBqiqMB0pF1Y=
=utBR
-----END PGP SIGNATURE-----


More information about the torqueusers mailing list