Dear All
I have now uploaded minutes from the meeting
(http://indico.cern.ch/conferenceDisplay.py?confId=19090) . If you are a
site admin and did not attend please take a look and let me know if you
have any questions.
High-level points:
* There are sites not correctly publishing data to APEL
* GridPP site admins should be aware of the SAM history review and
comment on the wiki about their site profile:
http://www.gridpp.ac.uk/wiki/SAM_availability:_May-July_2007. The August
SAM target is 85%.
* Several sites have now upgraded their worker nodes to SL4. The target
for GridPP sites is to have upgraded by the end of August - this is
especially the case for CMS sites intending to take part in CSA07. It is
understood that some sites will migrate when new hardware is
commissioned shortly after the August deadline. I am checking on the
question of using the SL3 software area and how big this area should be
going forward.
* Some sites have made the R27 upgrade. All sites will make the jump due
to the SL4 migration.
* There open tickets which the ROC is now chasing. Be aware that COD
tickets that do not receive a response get escalated within EGEE.
* All sites should contribute their thoughts on the ROC priorities to
raise with the EGEE Technical Coordination Group (TCG):
https://twiki.cern.ch/twiki/bin/view/EGEE/SA1_TCG. Send voting to
Philippa.
* A generic UKI ROC website is to be found here: http://www.ukiroc.eu.
What do you as users of it want to see developed here? For example ROC
processes and procedures, contact lists etc. Remember this is wider than
just GridPP....
* Everyone should now be aware of the GridPP policy regarding
stalled/inefficient jobs. Details are here:
http://www.gridpp.ac.uk/pmb/docs/GridPP-PMB-113-Inefficient_Jobs_v1.0.pd
f. There is in parallel with this a review of current queue lengths
being used.
* 2-3 sites have now affiliated with NGS. Such interoperation work is
how we will move towards a National Grid Infrastructure beyond GridPP.
Please speak to your Tier-2 coordinator about getting involved.
* Please consider supporting new VOs. From NGS we have ngs.ac.uk. There
is a request from a UK group: supernemo.vo.eu-egee.org. Also lookout for
your regional VOs which are being created.
* If you have ideas for the monitoring workshop/tutorial in London in
October please let me know.
* DPM sites should upgrade to the latest GridFTP server. Contact Greig
Cowan if you need more information.
Thanks,
Jeremy
> -----Original Message-----
> From: Testbed Support for GridPP member institutes [mailto:TB-
> [log in to unmask]] On Behalf Of Coles, J (Jeremy)
> Sent: 26 July 2007 10:10
> To: [log in to unmask]
> Subject: Plane VRVS room not Moon (was RE: Next UKI monthly operations
> meeting - Thursday 26th @ 10:30 in the Moon virtual room of VRVS)
>
> Dear All
>
> Please note that this meeting will now take place in the Plane VRVS
> room.
>
> Thank you to the GridPP sites (RHUL; Lancaster; Liverpool; RALPP) that
> have already responded to the request to comment on the SAM histories
> (http://www.gridpp.ac.uk/wiki/SAM_availability:_May-July_2007).
>
> Jeremy
>
>
> > -----Original Message-----
> > From: Testbed Support for GridPP member institutes [mailto:TB-
> > [log in to unmask]] On Behalf Of Coles, J (Jeremy)
> > Sent: 24 July 2007 18:11
> > To: [log in to unmask]
> > Subject: Next UKI monthly operations meeting - Thursday 26th @ 10:30
> in
> > the Moon virtual room of VRVS
> >
> > Dear All
> >
> > Our next monthly site meeting for UKI deployment and operations is
> this
> > Thursday. The agenda is here (but will evolve):
> > http://indico.cern.ch/conferenceDisplay.py?confId=19090.
> >
> > Ahead of the meeting please:
> >
> > 1) Record (for your quick reference) any issues your site may have
in
> > moving to SL4 by the end of August
> >
> > 2) Take a look at the GridPP document on stalled jobs:
> >
>
http://www.gridpp.ac.uk/pmb/docs/GridPP-PMB-113-Inefficient_Jobs_v1.0.pd
> > f. The intention is to go with this policy from 1st August to 31st
> > December and then review/modify in light of experience and further
> > input. This document is only a guideline for sites wishing to be
> > proactive in removing stalled jobs and does not require that sites
do
> > it!
> >
> > 3) Check the batch queue lengths that your site currently uses.
> >
> >
> > Finally, please could I encourage you to register with the video
> > conferencing system that will replace VRVS. It is called EVO. You
will
> > find all the information and links here: http://evo.caltech.edu/. We
> > will aim to try it for the September UKI meeting (today's DTEAM
> meeting
> > used it albeit with a couple of people not managing to get the
> software
> > installed/configured in time).
> >
> > Thanks,
> > Jeremy
|