Attending: Eric, Jay, Jeff P., Andrei, Marjorie and Jeff A.
Cluster Untilization: Usage has tapered off as we have reached Quark Matter.
Outages/Incidents: Various reports of intermittent slowness and bad connections, thought to probably be related to GPFS glitches (?).
Upcoming Downtime: 6/1 there is a NERSC network downtime so we will take a PDSF downtime to do the /home filesystem migration. It will be from 8am to 2pm and we will turn off io resource before it starts. A downtime for dayabay's eliza16 is also needed at some point but is not yet scheduled.
Procurements/New Hardware: The ALICE procurement has gone out. This is for about 400 cores plus ALICE storate. Will get rid of ~300 compute nodes and all of the current interactive nodes by this fall. ATLAS noted that they want to keep their share of the cluster about the same and will be purchasing at some point, too.
NERSC NX Server: PDSF NX server to be phased out and people should migrate to the NERSC server.
Grid Issues: No news from ATLAS. STAR has been having some issues copying output files from grid jobs run here back to RCF but it's thought that the problem is probably at RCF. Will do an SRM upgrade during the downtime on 6/1.
Jay reported on some work he's done on disk scanning. He created 2 lists of metadata, and it takes < 5 minutes on eliza1/2 but can take up to a day to do both lists on eliza18 (18M inodes).