Attending: Eric, Jay, Katie, Andrei, Jeff P., Marjorie
Utilization: Cluster has been completely full.
- Alicernc (alicesgm?) has been transferring on batch nodes.
Upcoming Downtimes: nothing until post-QM. Still plan to switch moe and upgrade GPFS on elizas.
Procurements/New Hardware: We will be getting ~1 rack's worth of (ALICE) comput nodes in the next month, ~500 cores. 2007 nodes are out of warranty - this is racks 9+10, ~40 noes or 300 cores. These will go away.
Training: Discussed the possiblity of Eric hosting a training session for PDSF users. A preference was expressed for web documentation instead.
New batch system: All is working except io resources. However, the additional benefits and feature might outweigh this deficiency and we will press on with the transition.
Grid Issues: Andrei reports success with ATLAS grid transfers - however, files were much smaller than when there were problems before.
- Discussed disk usage monitoring and Jay will look into generating data for the different groups.
- Discuss exit status 1 jobs: This is not understood at present...