On Tue, Jul 21, 2009 at 23:20, Ewan
MacMahon<[log in to unmask]> wrote:
>> -----Original Message-----
>> From: Testbed Support for GridPP member institutes [mailto:TB-
>>
>> On Tue, Jul 21, 2009 at 15:28, Ewan wrote:
>> > I'm a little concerned that when I raise the allowed number of
>> > running ones then we'll be running and finishing them faster than
>> > they're coming in.
>>
>> I would not worry - the pilot factory is doing the right thing:
>>
>> factory : INFO ANALY_OX: 772 activated jobs, 22 inactive
>> pilots queued (>= queue depth 10 * depth boost 2). No extra pilots
>> needed.
>>
>> Raise the bar and the system will rebalance quickly. It's dangerous to
>> send too many pilots to sites, they get stale.
>>
>> Anyway, you want to leave the system for a couple of hours to get good
>> stats at each measurement point.
>>
>
> Things have mostly worked as expected with more pilots coming in as more
>
> jobs are allowed to run, but it's falling behind now. I upped the number
> of running jobs from fifty to a hundred at about half past eight, but at
> the moment it's only running sixty-seven with none queued. Fresh pilots
> are still arriving, but apparently not fast enough.
OK, so I sent more pilots already! :-)
It's now aim to get 60 queued for Oxford (and most other sites).
Cheers
Graeme
--
Dr Graeme Stewart http://www.physics.gla.ac.uk/~graeme/
Department of Physics and Astronomy, University of Glasgow, Scotland
DEATH TO MEETINGS!
|