============================
Present: A. Forti, B. Davies, M. Doidge, A. Beresford, M. George, P. Trepka,
Sites report:
============
Lancaster:
==========
dcache is working now but it is still not published in the information system. It has been trsted but not yet with continuous requests. The might need to replace the srm.batch (or equivalent file) to cope with many queued requests. They would like to keep it separate for the SC3 and they don't want to publish it as production system. They are working on IP addresses to connect to ukLight. Lancaster university doesn't want their production network to be flooded. Matt will install the additional SC machines (LFC and FTS).
In addition Peter sent the report below.
I'm on holiday from tomorrow for one week so please accept my apologies, Matt and Brian should be attending.
Lancaster CPU resources have been running very stable since being online early April. We've had a couple of accounting issues with publishing old data but these have been resolved. As part of SC3 we will update our BDII to LCG-2_5_0 and install two new service machines (LFC, FTS).
About 50% (35TB) of our storage is online via SRM (dCache). This is currently connected via our university network, UKLight connection is being worked upon. Matt and Brian can give more details.
Liverpool:
==========
CE SE 1 WN UI still under test, NAT is still under work. Michael going away next week. Pawel has arrived last week.
Manchester:
===========
Order for new farm has been sent out. The old farm is up and that what can be seen on the SFT pages. It has two dcache pools on the 2 or the worker nodes which are the DELL tesst machines. To make it work reliably the srm-bohr0013.batch file has been changed. Now It works well when we don't have to change bring down one of the nodes. This needs investigation. Colin has started to work at the beginning of the month.
Sheffield:
==========
Farm running smoothly problems with BDII solved with the upgrade to new rpm and LHCB jobs now are running normally. Ordered 3.8 TB additional storage space.
Service Challenges:
===================
Brian reported from the SC3 workshop.
SC3 data transfers will be executed submitting proper
jobs to the system and using the Data Management structure. Sites in
SC3 are required to install LCG 2_5_0, LFC (LCG File Catalog), and FTS
(File Transfer Service). There was a discussion about the quality of
the machine for the LFC and what needs to be done in terms of
mirroring the disks, having a reliable power supply and frequency of backup.
There is also a request to have an experiment service machine per
experiment. The experiment service machine would run specific
experiment services nd possibly requires also opening accounts to
external people to run this services.
Tickets:
========
Alessandra plan to facilitate the adding and removing of sys admins who receive tickets inserting mailing lists rather than personal addresses in ggus didn't work due to a forced reply-from field. Asked for who is going to receive the tickets at sites so she can give a list to the CIC. In the future the names will be hopefully extracted from the GOC database which is Alessandra plan B. It has already been suggested to the CIC and to the project manager.
People per site:
Lancaster: Peter, Matt, Brian
Liverpool: Michael, Pawel
Manchester: Marc, Colin
Sheffield: Andy, Matt, Paul
Alessandra receives all the tickets.
GridPP meeting:
==============
NorthGrid talk is divided in 2 parts. Site reports for the past 6 months and then a general report,comments and feedback to LCG. Sites are asked as usual to report if there is anything they want to be said.
Next Meetings: 29 July 2005 on the phone.
Last modified Fri 24 June 2005 . View page history
Switch to HTTPS . Website Help . Print View . Built with GridSite 1.4.3