In my view in the CPU section of Steve's metric page:
http://pprc.qmul.ac.uk/~lloyd/gridpp/metrics.html
sites should be publishing the batch system fair-share value available
to ATLAS, which I suppose is what you call the guaranteed amount below.
Otherwise sites could get credit for offering excessive amounts of CPU
that ATLAS have in practice no hope of actually getting. If sites run
more then they will be credited with that in the CPU delivered sections.
Duncan
On 19/04/2011 16:40, Rob Fay wrote:
> I imagine it's similar for sites with
> access to a shared cluster (this would apply to the Computing Services
> cluster behind hepgrid3 at Liverpool). Do they publish the entire
> cluster, even though in practice they may well not be able to use all of
> it, or do they publish only the share they can guarantee, even though at
> other times they may well get more? Do they publish the maximum possible
> CPU share, or the minimum?
>
>
|