perfsonar development work

[Jochen Reinwand <>]

On Tuesday 29 April 2008 11:59, Nicolas Simar wrote:
> Jochen Reinwand wrote:
> > On Tuesday 29 April 2008 11:44, Nicolas Simar wrote:
> >>Jochen Reinwand wrote:
> >>>Hi all,
> >>>
> >>>One of the outcomes was that Martin wanted to put together a bunch of
> >>>possible solutions for the large data transfer problem.
> >>>Meanwhile, we (David, Verena and myself) are working on a specific
> >>>solution for our specific problem, that is based on one of the
> >>>suggestions Martin came up with in Zagreb. We are already working on it.
> >>
> >>What is the timescale we are talking about?
> >
> > Since the DFN wants to have it soon, it will hopefully be ready soon :)
> > But I think the most time consuming part is David's part. So perhaps he
> > can give a better estimate. We planned to be ready till the Technical
> > Workshop in June.
>
> Can you briefly describe the chosen solution, so we can check if it
> would work also for pSUI.

Very simple: Data is transferred via rsync in Perl binary format.
This also answers your questions: It will in no way work for pSUI, because 
the 
requirements are completely different.

AFAIK the problem with pSUI is the request used for filling the matrix. I'm 
working on extracting that sort of status overview information for Toby's 
applications. When this status information is available, it should be quite 
simple to make it available via a perfSONAR request. pSUI can use this 
request to fill the matrix and do similar things. That's one of the reasons 
why I insisted on a proper solutions for Toby's apps. The results (like the 
status information) will be usefull in a lot of other situations.

regards,
Jochen

-- 
Jochen Reinwand, DFN-Labor
Friedrich-Alexander-Universität Erlangen-Nürnberg
Regionales RechenZentrum Erlangen (RRZE)
Martensstraße 1, 91058 Erlangen, Germany
Tel. +49 9131 85-28689, -28800, Fax +49 9131 302941

www.win-labor.dfn.de