On 05/15/2012 10:01 AM, Mark Dixon wrote:
On Mon, 14 May 2012, iqtcub wrote:
...
I just tested with two virtual machines having 12 cores each. Same
result, if i don't use a PE, it works fine, however if i use the PE smp
i mentioned above(using $pe_slots as allocation_rule), it puts each job
in a different node.
I've also just tried using the binaries of OGS2011.11 with the same
result.
On the other hand, i've seen that in the other threads in the mailing
list, the people seem to use this option only when the jobs are of 1
core, which works perfectly.
So, any hints? Anyone has it working for more than one core?
...
Apologies in advance if I've got the wrong end of the stick.
It sounds like you're trying to get all jobs to fill your nodes in a
particular sequence, thereby allowing any usable "gaps" to be filled
by small jobs and keeping the maximum number of empty nodes handy in
case anyone wants to submit a big job. Is that right?
I don't know if I've missed anything myself but, in the past, I've
found that this (queue_sort_method seqno, job_load_adjustments NONE,
etc.) doesn't always work with PEs unless you specify a different
seq_no for each queue instance. You might want to give that a go.
All the best,
Mark
Hi Mark,
Yes, what we want to do is exactly what you said.
Regarding the option you mentioned, we want to do it using only one
queue, which is what we have at the moment(they don't like to limit the
runtime of the jobs).
Thanks
_______________________________________________
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
