Alex,

Can you run qhost and see if the memory value is also negative also??
If it is, then this bug was fixed in any release of OGS/GE.

Rayson



On Thu, Oct 18, 2012 at 6:53 PM, Alex Chekholko <[email protected]> wrote:
> Hi,
>
> Running Rocks 6, so whatever GE version is included there.
>
> h_vmem is set consumable and per job, 4G default:
>
> -bash-4.1$ qconf -sc |grep h_vmem
> h_vmem              h_vmem     MEMORY      <=    YES         JOB 4G       0
>
> each exec host has an h_vmem attribute set:
> -bash-4.1$ qconf -se scg3-0-11 |grep h_vmem
> complex_values        slots=16,h_vmem=60G
>
> pe "shm" is defined;
> -bash-4.1$ qconf -sp shm
> pe_name            shm
> slots              999
> user_lists         NONE
> xuser_lists        NONE
> start_proc_args    NONE
> stop_proc_args     NONE
> allocation_rule    $pe_slots
> control_slaves     FALSE
> job_is_first_task  TRUE
> urgency_slots      min
> accounting_summary FALSE
>
> A user is submitting a job with '-pe shm -l h_vmem=120G', and it's getting
> dispatched to a host that has h_vmem=60G defined.  How is that possible?
>
> And qstat reports negative h_vmem values, e.g.:
> -bash-4.1$ qstat -f -u '*' -F h_vmem
> ...
> [email protected]          BIP   0/16/16        12.12    lx26-amd64
>         hc:h_vmem=-80.000G
>   88866 0.50500 mCSRR57762 yxl          r     10/18/2012 09:17:21     1
>   89094 0.60500 G_ordermar elisaz       r     10/18/2012 15:03:39    15
> ...
>
> Maybe the sgeexecd needs to be cycled for the setting to take effect?  I can
> try that next.
>
> Regards,
> --
> Alex Chekholko [email protected]
> _______________________________________________
> users mailing list
> [email protected]
> https://gridengine.org/mailman/listinfo/users
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to