On 02/02/2015 11:37 AM, Elena Korolkova wrote:
> could you please remind me ce’s/queue names and I’ll check agis settings.
Here's our set-up in terms of "endpoints".
--------- TORQUE/MAUI CLUSTER (SINGLE CORE ONLY) ------------------
CLUSTER HEADNODE: hepgrid96.ph.liv.ac.uk
CREAM CEs submitting to TORQUE:
hepgrid5.ph.liv.ac.uk:8443/cream-pbs-long,
hepgrid6.ph.liv.ac.uk:8443/cream-pbs-long,
hepgrid97.ph.liv.ac.uk:8443/cream-pbs-long
--------- ARC/CONDOR CLUSTER (SINGLE CORE OR MULTI CORE ) ------------------
CLUSTER HEADNODE: hepgrid2.ph.liv.ac.uk
ARC CEs submitting to CONDOR:
hepgrid2.ph.liv.ac.uk:2811/nordugrid-Condor-grid
> Does single queue have the same name both in condor and in torque?
No.
There are three "single core" queues in TORQUE/MAUI cluster (one for
each CREAM CE that submits to it.)
There are no "multi core" queues in TORQUE/MAUI cluster.
There is one "single and/or multicore queue" in ARC/CONDOR cluster (one
ARC CE).
Cheers,
Steve
>
> Thanks
> Elena
>
> On 2 Feb 2015, at 11:30, Stephen Jones <[log in to unmask]> wrote:
>
>> Thanks Elena,
>>
>> BTW: we have two completely separate clusters. One, TORQUE/MAUI, can
>> run single core only. The other, ARC/CONDOR, can run single or multi-core.
>>
>> When droughts like this happen, the ARC/CONDOR lies
>> empty, while the TORQUE/MAUI remains full of single core.
>>
>> Is it difficult to keep the single cores coming to the ARC/CONDOR
>> (as they do to the TORQUE/MAUI) during these droughts?
>>
>> Cheers,
>>
>> Steve
>>
>>
>>
>> On 02/02/2015 11:17 AM, Elena Korolkova wrote:
>>> Hi Steve,
>>>
>>> see attached plot.
>>>
>>> Elena
>>>
>>>
>>>
>>> On 2 Feb 2015, at 11:06, Stephen Jones <[log in to unmask] <mailto:[log in to unmask]>> wrote:
>>>
>>>> Hi,
>>>>
>>>> I think there's an MCORE job drought on at the moment.
>>>>
>>>> See attached.
>>>>
>>>> I wonder if there's any chance of seeing some single core jobs over here to fill the void?
>>>>
>>>> Cheers,
>>>>
>>>> Steve
>>>>
>>>> <traffic.png>
|