Print

Print


Thanks Steve.
I would like to know how many of these environment vars are still needed.
Surely the VOs know where their software is on CVMFS without the site telling them?
Do we really have to tell all jobs the name of the SE as well as publishing it through the BDII per VO?
I hope some VO experts or other site admins can clear this up.
It would be really helpful to have a documented contract as Steve says.
Thanks,
Simon

________________________________
From: Testbed Support for GridPP member institutes <[log in to unmask]> on behalf of Stephen Jones <[log in to unmask]>
Sent: 19 September 2019 08:29
To: [log in to unmask] <[log in to unmask]>
Subject: Re: Script grid-env.sh

Hi Tony,

Since that script (and, perhaps, a c shell version of the same thing) is
in /etc/profile.d/, then all processes source it and get the resulting
environment. The script holds a few values that are useful to a VO's
jobs, usually

gridenv_set         "VO_ZEUS_SW_DIR" "/cvmfs/zeus.desy.de
<http://zeus.desy.de>"
gridenv_set         "VO_ZEUS_DEFAULT_SE" "se2.ppgrid1.rhul.ac.uk
<http://se2.ppgrid1.rhul.ac.uk>"
Etc.

>
> My understanding is that this script is probably provided by a glite
> package (not sure if this is correct).

If I remember correctly, the YAIM tool (now out of date) put this file
in place. The site-info.def file had all the values and they got written
in by YAIM, I think. Since YAIM is gone, we have to do it by hand now.

>
> The script below is in one of our old nodes, but .
> /usr/libexec/grid-env-funcs.sh is not located there in the new nodes.
>
> [root@node192 ~]# locate grid-env-funcs.sh.  ### node192 is one of our
> new installations with Centos 7
> /opt/glite/yaim/etc/clean-grid-env-funcs.sh
> /opt/glite/yaim/etc/grid-env-funcs.sh


Yes. They possibly came (a long time ago?) from YAIM originally. Or
someone made them by hand...


> Do you know what package provides this script and how this scripts can
> be configured automatically along with the node installation? Is this
> still using YAIM?

As far as I know, no package provides this; it is a "victim" of a
decision, quite a while ago, to de-support YAIM. Note: In my view, this
is a grid-wide issue related to the "contract" (i.e. the requirements)
that exists between the Grid and the VOs. We may need to tighten up the
way we deal with these "promises". I'm not sure they are well defined,
anymore, as YAIM played the role of a Standard. Without YAIM, I'm not
sure what the standard is. Hence I'm cc'ing Maarten, to ensure WLCG/EGI
are at least familiar with the issue. It's quite difficult to solve,
since it needs central co-ordination. I'm also cc'ing TB_SUPPORT in the
hope that other UK sites can explain how they manage the problem.

Anyway, since jobs refer to these values, I'm afraid it is your job to
write the file or similar.

Cheers,

Ste

--

Steve Jones                             [log in to unmask]
Grid System Administrator               office: 220
High Energy Physics Division            tel (int): 43396
Oliver Lodge Laboratory                 tel (ext): +44 (0)151 794 3396
University of Liverpool                 http://www.liv.ac.uk/physics/hep/

########################################################################

To unsubscribe from the TB-SUPPORT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=TB-SUPPORT&A=1

########################################################################

To unsubscribe from the TB-SUPPORT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=TB-SUPPORT&A=1