June 7
Attending: Eric, Jay and Katie from NERSC and users Andrei, Jeff A. and Jeff P.
Utilization: Cluster tends to be running at about 50% of capacity on average the past month.
Recent Downtimes/Outages: 6/1: SWO for most of the day: /home migration to new hardware was completed, updated BestMan on pdsfdtn1. This upgrade will hopefully address some timeout bug that had affected STAR transfers.
Upcoming Downtimes: No clusterwide outages planned but eliza3 needs a reboot. This shouldn't require a downtime but might degrade performance for a while.
Question: Are we on new /common yet? Answer: No, has to do with investigations of using moab instead of maui with torque.
Procurements: Dell has been difficult with the Alice procurements because they want some changes which are difficult for us to make so it's been delayed a bit.
Project Accounts: Eric reported that a working project account is in place an is being test. Will also need to test logging w/security to be sure we record what we need.
Torque: Eric reports that he's working on webpages and a pbsusers script, similar to the old sgeusers script is available.
Grid Transfers: Andrei reports that ATLAS transfers are still working, at least for small files, but there is a problem with deleting files that therer is a ticket about. He say this isn't realy high priority but they will want to get it working.
Andrei reports reading on a new cvmfs mailing list that a cache size of about 50GB works well and that is what we had decided months ago.