Testbed Support for GridPP member institutes
> [mailto:[log in to unmask]] On Behalf Of Mona Aggarwal said:
> To solve the above problem the RB scheduling should be design
> to handle
> Advance Reservation and to priotise job submission by the user
> deadlines etc. Thus individual (VOs) don't have to design extra
> frameworks.
The problem is that the whole history of the grid projects is that the
middleware hasn't delivered what the experiments need so they have gone
away and invented their own. Indeed the ARDA report in 2003, which lead
to the glite project, explicitly asked for this kind of thing, but four
years later it still hasn't been delivered - and even when we do get it
(glite WMS + CREAM) it isn't obvious that it will satisfy the
requirements.
> Then we should fix the SITES and WN's nodes so that we can
> provide more resources for the GRID.
Again, we've been trying to fix the sites for many years, and some
progress has been made - but it's far from good enough. As it happens
atlas is having a workshop on its job submission framework right now:
http://indico.cern.ch/conferenceDisplay.py?confId=23614
If you look in the talk by Simone Campana (last slide) you can see that
the current job efficiency for EGEE is 55.8% ... hence atlas is now
moving to the Panda pilot framework.
Stephen
|