[gridengine users] Grid efficiency from the point of view of memory usage

Reuti reuti at staff.uni-marburg.de
Thu Jul 21 12:33:02 UTC 2016


Hi,

> Am 21.07.2016 um 13:59 schrieb <sudha.penmetsa at wipro.com> <sudha.penmetsa at wipro.com>:
> 
> Hi,
>  
> We see that most of the cores in our grid queue are idle but my jobs are not getting cores because there is insufficient h_vmem in the queue as the memory is already used by other jobs
>  
> Would it be possible to somehow take memory requirements into account in the scheduling?

For the scheduling it is already being taken into account, therefore they are not scheduled to nodes without enough resources.

But to avoid starvation of your jobs: did you make h_vmem a consumable complex, attached a certain amount to each exechost `qconf -me ...` and submit your jobs with `qsub -R y` to avoid this effect?

Note: max_reservation in `qcoonf -msconf` must be set to the job count you expect that reservations are necessary for.

Then h_vmem will be reserved and at one point your jobs will start essentially.

-- Reuti


>  
> Regards,
> Sudha
>  
> The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email. www.wipro.com _______________________________________________
> users mailing list
> users at gridengine.org
> https://gridengine.org/mailman/listinfo/users





More information about the users mailing list