[Wien] Wien 2K Startup

[Herbert Fruchtl]

Most applications of SGE have little to do with the Grid. The Grid is usually
not about parallel computing, but about resource distribution on a higher
level. You would run a complete Wien2k calculation on one of several sites that
have Wien2k installed, but which site this is, which OS, user name, accounting
mechanism, access to licenses etc., is handled by the Grid middleware.
Multi-stage jobs may be scheduled to have their parts done on different "Grid
nodes", with intermediate data transfered between them. In the case of Wien2k
it probably doesn't make sense to split it up, except for the initial,
interactive part, which should be done locally.

I must say that I am grateful for Maliks questions and the answers he got. For
somebody used to running calculations in various batch environments where
interactive access is difficult and discouraged, the "normal" way of running
Wien2k takes some getting used to.

  Herbert

Quoting Torsten Andersen <thor at physik.uni-kl.de>:

> And I thought that grid environments provided functionality similar to
> those of cluster-queue-systems, such as SGE. There, environment
> variables are set that tell you about the system, and usually it can be
> done detailed enough to piece together a .machines file on-the-fly.
>
> Well, then I can't provide any more help... I do not have experience
> with SETI-like distributed systems...
>
> How would you treat systems with different performance in a
> geographically distributed system? There will have to be a variable
> telling you how to distribute the calculation efficiently.
>
> Best regards,
> TOrsten Andersen.
>
> Stefaan Cottenier wrote:
> >>It would have been more efficient if you had stated your intention at
> >>the beginning of your questions.
> >
> >
> > ...what is what he actually did in his very first post about 2 weeks
> > ago, and which was the reason I had more than usual patience ;-).
> >
> >
> >>As far as I know we have already a
> >>solution for your problem  called "wien2grid" ....
> >
> >
> > My knowledge about grid environments is very limited, but from the
> > stories I hear there is no universal standard yet, and every grid site
> > uses slightly or largely different middleware flavours. Could that be
> > a good motivation for the simultaneous development of different wien2k
> > grid solutions? Or do I badly miss the point?
> >
> > Stefaan
> >
> >
> > Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
>
> --
> Dr. Torsten Andersen        TA-web: http://deep.at/myspace/
> AG Hübner, Department of Physics, Kaiserslautern University
> http://cmt.physik.uni-kl.de    http://www.physik.uni-kl.de/
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>


-- 
Herbert Fruchtl
EaStCHEM Fellow
School of Chemistry
University of St Andrews


------------------------------------------------------------------
University of St Andrews Webmail: https://webmail.st-andrews.ac.uk