On Thu, 1 Nov 2007, Alessandra Forti wrote:
> Hi,
>
> > So, it is generally recognised that this situation has arisen due to the
> > interpretation of one user. That takes us back to the question Paul
> > raised earlier in the thread and I'm sure that will feature in the
> > follow up to this matter.
>
> If I belong to a VO and I run code that has nothing to do with that VO
> whatsoever or whose goal has nothing to do with the VO goals whatsoever
> I would say that's abuse of resources. The interpretation should be as
> simple as that.
Then the question becomes, who makes the interpretation? :(
>
> The question that is more difficult to answer is how do I know that the
> user is not running something outside this boundaries. What if Heinz had
> been adapting algorithms to breaking proteins rather than factorizing
> numbers?
What if some user submitted boinc clients to run say (protein)
Folding@home ? That would seem to satisfy the biomed AUP, yet it would
reduce the resources available for normal biomed jobs, and the scientific
benefit goes to some other (real) organisation. But then maybe it's good
for existing "legitimate" work to get independent confirmation of some
results sooner...
I wouldn't pretend to be remotely qualified to decide on something like
that - it would have to be the VO managers, who (presumably) understand
the science as well as the Grid issues involved.
My "relaxed" policy is partly laziness, but also I think a pragmatic one -
there are things I really don't want to get dragged into, and actively
policing VO AUPs is definitely one of them!
Of course, it does assume that the VO managers a priori are properly
informed about what their users are up to (something not clear in this
instance). And the VO managers can always ask sites to check on what
particular jobs are up to ...
Thanks
Henry
--
Dr. Henry Nebrensky [log in to unmask]
http://people.brunel.ac.uk/~eesrjjn
"The opossum is a very sophisticated animal.
It doesn't even get up until 5 or 6 p.m."
|