Hi
To clarify on
> There is a proposal that RAL runs a SAM for PPS so this will make sure
> basic tests are checked.
SAM PPS *will* be run by PPS-CYFRONET and PPS-RAL from 8th June (this is
probably the earliest it could be done and may take several weeks
longer). However, using the service and results is a ROC responsibility
and this will not change - the basic tests are already hosted elsewhere:
https://monitoring.egee.man.poznan.pl/admin2/ [site type PPS; node type
gLite-CE]. PPS-RAL
https://lcg-sam.cern.ch:8443/sam-pps/sam.py?funct=ShowHistory&sensors=gC
E&vo=ops&nodename=pps-ce-277.gridpp.rl.ac.uk and UKI-LT2-HEP-PPS
https://lcg-sam.cern.ch:8443/sam-pps/sam.py?funct=ShowHistory&sensors=gC
E&vo=ops&nodename=imalaydee.hep.ph.ic.ac.uk show the results for the
gLite-CE sites.
https://lcg-sam.cern.ch:8443/sam-pps/sam.py?funct=ShowHistory&sensors=CE
&vo=ops&nodename=epbf005.ph.bham.ac.uk gives the UKI-SOUTHGRID-BHAM-PPS
results. The latter does not show up the current installation problems
(Yves may wish to comment on how well the SAM results can reflect
problems being encountered).
With regard to SAM/GridView in general there is a meeting on Monday the
slides for which may be of interest and can be found here:
http://indico.cern.ch/conferenceDisplay.py?confId=16476
Jeremy
> -----Original Message-----
> From: Testbed Support for GridPP member institutes [mailto:TB-
> [log in to unmask]] On Behalf Of Gordon, JC (John)
> Sent: 18 May 2007 11:37
> To: [log in to unmask]
> Subject: Re: SAME tests run in job wrapper
>
> Even if there is a release every week, I think the period for testing
> individual items on the PPS is two weeks. From a discussion I had the
> other day, the concern at CERN was more that the VOs didn't confirm
that
> patches still worked for them.
>
> There is a proposal that RAL runs a SAM for PPS so this will make sure
> basic tests are checked.
>
> John
> > -----Original Message-----
> > From: Testbed Support for GridPP member institutes
> > [mailto:[log in to unmask]] On Behalf Of Peter Gronbech
> > Sent: 18 May 2007 11:17
> > To: [log in to unmask]
> > Subject: Re: SAME tests run in job wrapper
> >
> > Yves is of the opinion that CERN are no longer doing testing
> > of the updates and are relying on the PPS to do the testing.
> > If the PPS sites are busy and do not manage to test properly
> > (or at all) then the release comes out anyway. The pressure
> > seems to be to release every week, which in my opinion is too often.
> > Given the rate at which new bugs are apearing in the releases
> > it seems its better to have one site be a test site.
> > I think we have been here before!
> >
> > The new changes to YAIM mean that if you actually read the
> > documentation and run YAIM in the new way it fails to
> > complete and requires a CTRL c to finish it. If you continue
> > to run yaim in the way we always have done that particular
> > bug does not show its self.
> >
> > Other problems seem to be that when yaim is run on every
> > worker node it tries to do a recursive chown on the software
> > directory (which is usually on the ce). This take a lot of
> > time and is repeated by every worker node!
> >
> > Pete
|