On Wed, Nov 02, 2005 at 03:08:57PM +0100 or thereabouts, David Smith wrote:
> On Wed, 2 Nov 2005, Steve Traylen wrote:
>
> > I have an idea that people were talking about passing the
> > the value LCG_GFAL_INFOSYS through the RB submitted job
> > so the job itself used this value and so is able to see
> > the same view of the world.
> >
> > Is this now possible?
>
> Hi Steve,
>
> The broker does not automatically set LCG_GFAL_INFOSYS, but it can be
> specified in the JDL in the same way as any environment variable.
> (Although there is some issue then that csh shells may trigger souring of
> profile.d/ scripts which reset it).
>
> Last year we did try setting LCG_GFAL_INFOSYS automatically, for a few
> weeks, but at the time this caused problems with the loading of the BDII -
> which by definition was also the one that the RB was using for matchmaking
> queries.
>
> There aren't any plans to do anything about this in the near future - do
> you have a particular reason for asking now?
Thanks, it was trying to improve the results at
https://lcg-sft.cern.ch/sft-uncert/lastreport.cgi
where the RB/BDII used for these includes non certified sites. The
rep-man tests are clearly doomed in the current situation.
So adding
Environment = {"LCG_GFAL_INFOSYS=lcgbdii02.gridpp.rl.ac.uk:2170"} ;
is the way forward. I'll give it a whirl though really this is a request for
the people running.
https://monitoring.egee.man.poznan.pl/admin2/
if they are watching.
Steve
>
> Yours,
> David
>
> --
> -------------------------------------------------------------------------
> David Smith e-mail: [log in to unmask] tel: +41 22 76 74462
> Address: D. Smith, CERN G06610, Bat 28 R-007, 1211 Geneva 23, Switzerland
> -------------------------------------------------------------------------
--
Steve Traylen
[log in to unmask]
http://www.gridpp.ac.uk/
|