Testbed Support for GridPP member institutes
> [mailto:[log in to unmask]] On Behalf Of Dr D J Colling said:
> I don't think that this is the reason. I am not sure how the
> gstat gets its
> figures but we certain have all the major RB (with the exception of
> NIKHEF).
Not all jobs go through RBs though - as far as I know atlas still submit
jobs directly with condor-g, and there may be local jobs. I assume that
gstat sums all the RunningJob counts for the CEs, and those should be
read directly from the batch systems. There might also be some timing
differences. Also the RBs don't always see the Done events correctly so
they can report jobs as running which aren't actually there.
Stephen
|