NERSCPowering Scientific Discovery Since 1974

October 25

Attending:  Eric Marjorie Lisa Elizabeth

Cluster Utilization:  Running at capacity with over 1800 cores in services.  Some nodes are out of warranty, though so this number will gradually shrink as those nodes will not get fixed when they fail.

Outages:  None.

Upcoming Downtimes:  There will be an NGF outage in December but the date is still being nogotiated.

Procurements/New Hardware:  20 new compute nodes are in service now as are 2 new xrootd servers for ALICE.  The eliza16 expansion for dayabay is done.  No more procurements planned for this year.

Job Benchmarks:  Eric showed and discussed the job benchmarks pages.  Notable was how much faster the new nodes seem to be and this is not understood yet and more tests are planned.

Grid Issues:  There was some confusion over whether we were running a BDII service but it turns out we are running it.

SL6:  There is chos work that needs to be done but it could probably be in place in January.

Project Accounts:  Eric showed the project accounts webpage that describes how to use project accounts.

Topics from users:  Marjorie asked about batch system migration and the decision to stick with supported SGE was announced.  This is in part because JGI will be using it as well.  Torque is not completely dead but a migration to torque is on hold for now.