[gridengine users] "Packing" jobs on nodes

James Gladden gladden at chem.washington.edu
Fri May 27 19:09:01 UTC 2011


I have no such explicit resource quota configuration (as describe below) 
on my systems, yet no host slot over-subscription occurs.  I just tried 
the experiment on an 8 slot node with two queues assigned.  When I 
submit jobs to the specific queue instances associated with that 
execution host, the first 8 jobs I submit get dispatched to the node 
while the remainder (from either of the two queues) wait in the "qw" state.

The only thing I have done to facilitate this behavior is to set the 
value of the consumable "slots" resource for each execution host to 8 
(which happens to be the number of cores on each execution host).  
Presumably if I had wanted to allow over-subscription, or utilize 
hyper-threading, I could set the value to something larger.

My conclusion is that enforcing the "slots" resource limit on hosts is 
the default behavior for SGE.  Has anyone actually observed different 
behavior?

James Gladden

On 5/18/2011 1:48 PM, Dave Love wrote:
> "Hung-ShengTsao (Lao Tsao) Ph.D."<laotsao at gmail.com>  writes:
>
>> as a side remark
>> since you configure 3 queues for each host , in order not to over subscribe the
>> cpu(h_slots)
>> you will need to define h_slots as consumable complex
>> see this paper http://www.sun.com/blueprints/0607/820-1695.pdf
> ??
> All you need to avoid over-subscription, whatever the queue defs, is
>
> $ qconf -srqs host-slots
> {
>     name         host-slots
>     description  "restrict slots to core count"
>     enabled      TRUE
>     limit        hosts {*} to slots=$num_proc
> }
>
> assumong the default slots definition
>



More information about the users mailing list