MeetingMinutes08Jan2007

• Present
  • MAH, CG, BTK, JB, RF, SJM
• Apologies
  • PA, PT

LCG cluster report

1. Status of infrastructure. OK, no problems with aircon etc.
2. Problems: Hepgrid5 (the SE) has developed an access problem that is being investigated. This machine is needed for CG to install ATLAS v12.04 on the cluster.
3. Dcache and SE Status: Dcache is under test by PT. There is not much space left on the SE (~0.7 TB) to store ATLAS data. The meeting requested an audit of what disk space is available for HEP users to use in the cluster room, and a ~6 month forward look to estimate how much ATLAS data we have to store here. Action JB/RF/CG.
4. Internal network issues: FORCE10/DELL switch connectivity. One cabinet has an intermittent fault that causes the cluster to go off line for short periods. It was suggested that the DELL switch be changed. Possible upgrades to the FORCE10 switch software are being investigated.
5. Plans: upgrades/mods whatever. PA has agreed that memory can be bought to repair ~20 nodes. JB and RF are doing a complete audit of all the outstanding and part-finished tasks left over on change of System Administrators.
6. ATLAS DQ2 catalogue software: CG has set this up on hepstore and circulated a note to ATLAS users about how to use it. It works fine for him but other users are requested to test it.

Non-LCG cluster report

1. Number of racks in use: 3 racks are used at present, but this will increase to ~5 in the near future. As this facility will be heavily used in the coming months, any spare DELL nodes, such as any repaired ones, should be added to this facility.
2. Status of users' software: Front end machines for Cockcroft and T2K users are being installed. The system is running at full capacity with ATLAS MC.
3. Statistics (it will be useful to record the use of this facility): For the time being, users are requested to keep records of any significant use of this facility as there is no automatic system yet in place to record statistics.

Trash/CDF and SAM

1. Linking SAM to LCG nodes. The SAM station has been fully updated by VB at UCL. It needs linking to LCG nodes. It was agreed that two racks on LCG be linked.

Network issues

1. Ok, no problems.

Trash/BaBar Cluster

1. ATLAS wishes to use this, and currently needs SLinux3. There is a plan for ATLAS to change to SLinux4 in the near future. It was agreed at the meeting that the multi-node (~10) interactive facility should use part of the Trash/BaBar cluster so that all the DELL nodes can be exploited on the main clusters.

Desktop upgrades

1. Report on Desktop upgrade progress (RF/JB). This was successful, and only a few minor tweaks are needed.

Plans

1. Documentation. Over the coming months JB and RF will produce much documentation about all aspects of the system that will be installed on the Twiki.
2. Clean Room PC upgrades. Some of the PCs in the clean room are legacy machines running old versions of windows etc to support specialist interfaces. It's planned to review this situation and try to upgrade both the OS and hardware to make these machines easier to maintain.

Any Other Business

1. None.

Date of next meeting