Is this working at any sites, I mean are there lhcbprd jobs rather than
lhcbsgm jobs actually running? I am finding that I have lhcbsgm jobs
queuing which fail immediately.
Duncan
On Wed, 2006-09-20 at 14:06 +0100, Yves Coppens wrote:
> Hi,
>
> There has been a follow up on this in "the New YAIM groups.conf file for
> LHCb (Errata Corrige)" thread in ROLLOUT.
>
> I've copied the relevant message below.
>
> Yves
>
> <--
> Sorry for flooding your mail boxes,
>
> as Cristina correctly pointed out the lines to add on the
> groups.conf files (and that we had test at that time here at CERN too)
>
> were:
>
> "/VO=lhcb/GROUP=/lhcb/sgm":::sgm:
> "/VO=lhcb/GROUP=/lhcb/lcgprod":::prd:
>
> and not
> "/VO=lhcb/GROUP=/lhcb/sgm/Role=NULL/Capability=NULL\" lhcbsgm:
> "/VO=lhcb/GROUP=/lhcb/lcgprod/Role=NULL/Capability=NULL\" lhcbprd:
>
> as originally stated (it was a typo)...
>
> This is exactly what LHCb would like to have at each site.
>
> Thanks again for your support and comprehension
>
> R.
> -->
>
> On Mon, 18 Sep 2006, Alessandra Forti wrote:
>
> > Hello,
> >
> > Lhcb is having user mapping problems (see below). It explains why they
> > keep on using unwillingly lhcbsgm for production.
> >
> > Below there's the explanation and recipe to hopefully correct this
> > behaviour. Could you please apply it so they can have a uniform mapping?
> >
> > cheers
> > alessandra
> >
> >
> > -------- Original Message --------
> > Subject: New YAIM groups.conf file for LHCb
> > Date: Mon, 18 Sep 2006 14:50:33 +0200 (MEST)
> > From: EGEE BROADCAST <[log in to unmask]>
> > To: [log in to unmask], [log in to unmask],
> > [log in to unmask], [log in to unmask], [log in to unmask],
> > [log in to unmask], [log in to unmask],
> > [log in to unmask], [log in to unmask],
> > [log in to unmask]
> >
> >
> > ------------------------------------------------------------------------------------
> > Publication from : Roberto Santinelli 7735 <[log in to unmask]>
> > (CERN)
> > This mail has been sent using the broadcasting tool available at
> > http://cic.in2p3.fr
> > ------------------------------------------------------------------------------------
> >
> > Hello,
> >
> > according the LHCb VO-card updated on the CIC portal, LHCb do use VOMS
> > groups instead of Roles (as Atlas and CMS do) for mapping users to the
> > special lhcbprod or lhcbsgm local accounts. The currently deployed
> > default configuration of YAIM (groups.conf) doesn;t support this model
> > but rather:
> > [root@ce101 root]# grep lhcb /opt/edg/etc/lcmaps/gridmapfile
> > \"/VO=lhcb/GROUP=/lhcb/ROLE=lcgadmin/Capability=NULL\" lhcbsgm
> > \"/VO=lhcb/GROUP=/lhcb/ROLE=lcgadmin\" lhcbsgm
> > \"/VO=lhcb/GROUP=/lhcb/ROLE=production/Capability=NULL\" lhcbprd
> > \"/VO=lhcb/GROUP=/lhcb/ROLE=production\" lhcbprd
> > \"/VO=lhcb/GROUP=/lhcb/Role=NULL/Capability=NULL\" .lhcb
> > \"/VO=lhcb/GROUP=/lhcb\" .lhcb
> >
> > A correct behavior for LHCb would imply these further lines in the
> > LCMAPS gridmap file:
> > \"/VO=lhcb/GROUP=/lhcb/sgm/Role=NULL/Capability=NULL\" lhcbsgm:
> > \"/VO=lhcb/GROUP=/lhcb/lcgprod/Role=NULL/Capability=NULL\" lhcbprd:
> >
> > that may be easily achieved by adding to the groups.conf file before the
> > other lines for LHCb the following two lines :
> >
> > \"/VO=lhcb/GROUP=/lhcb/sgm/ROLE=NULL\":::sgm:
> > \"/VO=lhcb/GROUP=/lhcb/lcgprod/ROLE=NULL\":::prd:
> >
> > And then rerun at least YAIM\'s config_mkgridmap:
> > /opt/glite/yaim/scripts/run_function your_site_info.def config_mkgridmap
> >
> >
> >
> >
--
Duncan Rand, School of Engineering and Design,
Brunel University, Uxbridge, UK
Email: [log in to unmask] Tel. +44 1895 266804
|