Dear All,
As you may know we collect various pieces of performance information about
the LCG. Now, we were interested to see how to interpret this
information and we thought that a model of the LCG might be useful in
trying to understand which numbers are actually important. So this model
would have to be tuneable so that we could add the timings and job retries
etc.
It occurred to me that a few years ago lots of people seemed to be writing
computer models of the grid (then edg). I was wondering if any group had
continued with this activity and had a model that we could use in the way
that I described? Clearly, there may be additions that we would have to
make and bits of tweaking etc, however it would seem pointless us
developing a model if there is already something out there.
So if anybody does know of modelling activity then we would be grateful if
you would get in touch.
All the best,
david
|