[gridengine users] Making the fair-share policy/scheduler algorithm "more fair"

Mark Dixon m.c.dixon at leeds.ac.uk
Wed May 1 10:02:20 UTC 2013


On Tue, 30 Apr 2013, Dave Love wrote:

> Mark Dixon <m.c.dixon at leeds.ac.uk> writes:
>
>> We use the share tree here, rather than the functional policy, so this
>> might not be applicable.
>>
>> By default, the "usage" of a job is wholly based on slots*seconds.
>
> I think it's only (effectively) slots if you use
> sharetree_reserved_usage, as we do.  I'm fairly sure man and qconf are
> right in saying CPU seconds.

I deliberately glossed over the definitions, to concentrate on more basic 
things. Particularly as the OP is using the functional policy and so 
presumably SHARETREE_RESERVED_USAGE doesn't do anything :)

By default, it's the integral of used-slots*cpu-seconds for cpu and (I 
think) used-gigabytes*slots*cpu-seconds for memory.

If you have execd_params SHARETREE_RESERVED_USAGE=true, it's the far more 
sensible requested-slots*wallclock-seconds for cpu and 
requested-gigabytes*slots*wallclock-seconds for memory (i.e. charged for 
the resources you've prevented other jobs from using).

Mark
-- 
-----------------------------------------------------------------
Mark Dixon                       Email    : m.c.dixon at leeds.ac.uk
HPC/Grid Systems Support         Tel (int): 35429
Information Systems Services     Tel (ext): +44(0)113 343 5429
University of Leeds, LS2 9JT, UK
-----------------------------------------------------------------



More information about the users mailing list