[Rivet] Running Rivet on the grid

[Frank Siegert]

Andy Buckley, Wednesday 15 October 2008:
> > Do you think it is feasible to link AGILe at runtime (if found in
> > LD_LIBRARY_PATH) instead of compile time? Since we have identified
> > all relevant places with an "#ifdef HAVE_AGILE" already, this might
> > not be so hard? Of course, during compilation one would need the
> > AGILe headers anyways. If nobody has a better suggestion, I could
> > look into this. Another option would be providing two Rivet tarballs,
> > one compiled with AGILe, and one without.
>
> The new interfaces that I've been working on a bit in the last week
> completely do away with using AGILe programmatically. Instead you run
> AGILe to dump HepMC events to a FIFO (or an unnamed pipe, once I've
> worked out how to handle arbitrary file descriptors) and then run
> rivet's interface to read from the same pipe. If sockets have something
> to offer, then I'm happy to add support for them, too. There's no Rivet
> dependency on access to AGILe, and other generators like Sherpa can do
> it without needing to use AGILe at all (although this then loses the
> uniform way to pass parameters).

Cool, I didn't know that there was any work on this in AGILe yet.

> >  - Rivet
> >    + libgsl.so link (from what i have noticed, this is missing on
> > SL4, but might only be needed during compilation?)
>
> It's used in the sphericity calculator to diagonalise the momentum
> tensor.

I meant, that on SL4 the 
symlink /usr/lib/libgsl.so -> /usr/lib/libgsl.so.0.6.0 is missing, and 
Rivet compilation fails because of that. Or is that only on Durham's Grid 
UI machines ("swainby" and "ui") the case?
And is it only needed at compile time, or also at run time?

> Yes. For now I would use the AGILe version of Rivet only, since it
> keeps things simple, and anyway the download size of the AGILe library
> is pretty small: about 1MB.

Yes, of course. I was not really worried about the size of the AGILe 
library itself, but mainly with the growing size when you add all 
generators in at the same time.

But these new interfaces will make things much easier (at least from the 
validator point of view) :)

Cheers,
Frank