Hi All, 

UGE 8.2.0 (which is in Beta now) also provides h_rss (resident set size) and 
Univa Grid Engine will now kill jobs (if you set h_rss) if they exceed the 
Resident Set Size.

and if you use cgroups - you can configure 'cgroups assistance' on killing jobs 
which means you no longer need to turn on the supplementary group id to track 
processes of a job.

Regards, 

Bill.

On Jun 6, 2014, at 2:36 PM, Skylar Thompson <[email protected]> wrote:

> We've started using cgroups and m_mem_free to accomplish this, using UGE
> 8.1.7.
> 
> On Fri, Jun 06, 2014 at 02:30:46PM -0400, Michael Stauffer wrote:
>> OGS/GE 2011.11p1
>> 
>> Hi,
>> 
>> Is there an alternative to h_vmem that checks resident memory rather than
>> virtual? I'd like a consumable that kills a job when it oversteps its
>> requested memory quota, so h_vmem is great. But I'm having trouble with
>> Matlab (and maybe other apps, haven't checked fully yet) which at startup,
>> allocates a somewhat-variable large amount of VIRT memory (~4G) but little
>> RES memory (~300M) (according to 'top').
>> 
>> Matlab's java environment is allocating most of the virtual memory, but
>> they can't tell me a way to limit this. In any case this large VIRT value
>> makes it impossible to set a smaller default h_vmem value because Matlab
>> won't launch.
>> 
>> Thanks for any thoughts.
>> 
>> -M
> 
>> _______________________________________________
>> users mailing list
>> [email protected]
>> https://gridengine.org/mailman/listinfo/users
> 
> 
> -- 
> -- Skylar Thompson ([email protected])
> -- Genome Sciences Department, System Administrator
> -- Foege Building S046, (206)-685-7354
> -- University of Washington School of Medicine
> _______________________________________________
> users mailing list
> [email protected]
> https://gridengine.org/mailman/listinfo/users

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to