London Tier 2 Technical Group.

Meeting 1.
6 May 2004, 1:00pm
Rm 532 Blackett Lab
Imperial College London

Attending:
Steven Newhouse (LeSc)
Alex Martin (QMUL)
Ben Waugh (UCL)
Paul Kyberd (Brunel)
Grigori Rybine (RHUL)
Sukhbir Johal (RHUL)
David Colling (Imperial)
Owen Maroney (Imperial)

1.1 Does there exists documentation on the Service Level required of an LCG-2 site.  In particular:
 - What is the timescale on which a site needs to upgrade?
 - What are the circumstances in which a site may be removed from the Core zone?

Action on OM: find if such documentation exist and if so, disseminate and provide link on London Tier 2 webpage.

1.2 Does there exits documentation on Experiments requirements for LCG-2 site.
Action on OM.  find if such documentation exist and if so, disseminate and provide link on London Tier 2 webpage.

1.3 GOC Monitoring.  Need to understand what the GOC monitoring actually does, so that if a site fails monitoring will be able to reproduce the failure mode.
Action on OM.  Contact GOC, via TB-Support mailing list.

1.4 Minimal Installation on WN.  Recipe is required for "minimal installation" on WN on existing PBS farms.  Will be based upon the previous EDG installation and the LCG Manual RPM installation guid.
Action on OM.  Investigate existing documentation, seek more detailed recipe from LCG.

1.5 System Support.  While user support is provided by GridKA and basic monitoring by GOC, there is no specific centre to provide system support.  It is believed that the Tier 2 sites will get their system support from their (Tier 1) Regional Centre.  In this case it will be RAL, and using the TB-Support Mailing List.
Action on OM. Confirm this is the correct support mechanism.

1.6  Source RPMs.  Is it possible to obtain source rpms for LCG installation?
Action on OM.  Find out.

1.7  VO Monitoring.  LeSc has software for monitoring usage of Grid farm by VO.  If possible this could be deployed at other sites, to allow reporting of usage of farms by VO.
Action on LeSc.  Develop software for deployment at other grid sites.

1.8 Website.  OM announced the existence of the London Tier 2 website.
 - All to supply DN to OM to provide suitable read/write access
 - Webpage for each participating site
 - FAQ page gathering together all useful information in one page (this may be made part of the overall UK TB-Support pages)
Actions on OM, all.

1.9 Mailing lists.
 - Update the "comprehensive" LT2 mailing list.  
 - Create "technical group" mailing list.
Actions: OM to supply new emails to DC.
 DC to update "comprehensive" mailing list.
 DC to make OM adiminstrator of "comprehensive" mailing list.
 DC/OM to create "technical group" ML.

1.10  Firewall and iptables.  Frontend nodes are often outside the site firewall and do not have automatic iptables configuration.  List of required inbound ports is known.
Actions on OM:
 - create standard iptables configuration for frontend nodes.
 - disseminate this, and firewall ports document, on FAQ webpage.

1.11  Private addresses and Outbound WN access.  Use of iptables/NAT to allow private IP address WN to get outbound access.
 - OM to discuss with AM with view to providing recipe for Brunel and RHUL.
 - Definition of outbound WN needs to be provided to Brunel.

1.12 GridICE.  Create GridICE website for London Tier 2 sites.
Action on OM.

Next meeting?