[slurm-users] Resources are free but Job is not getting scheduled.

[navin srivastava]

i missed to add the scheduling parameter.


SchedulerType=sched/builtin
#SchedulerParameters=enable_user_top
SelectType=select/cons_res
#SelectTypeParameters=CR_Core_Memory
SelectTypeParameters=CR_Core

# JOB PRIORITY
PriorityType=priority/multifactor
PriorityDecayHalfLife=2
PriorityUsageResetPeriod=DAILY
PriorityWeightFairshare=500000
PriorityFlags=FAIR_TREE

could you please also suggest here if the scheduling policy is fairshare
then still it will consider the priority over the partition?

Regards
Navin.


On Sat, Apr 4, 2020 at 8:34 PM navin srivastava <navin.altair at gmail.com>
wrote:

> Hi Team,
>
> I am facing one issue in my environment. our slurm version is 17.11.x
>
> My question is i have 2 partition:
>
> Queue A  with  node1 and node2  with Priority=1000 shared=yes
> Queue B with node1 and node2  with priority=100. shared =yes
>
> Problem is when job from A partition is running then the job from
> partition B is not going through eventhough the cpu is available on node1
> and node2.
>
> it is only accepting the job from partition A but not from B.
> Oversubscription=No tells job will run from both the partition then it
> should allow.
>
> Any suggestion.
>
> Regards
> Navin.
>
>
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20200404/808a0ee5/attachment.htm>