Meeting Notes for 27 Feb

General

• Stress test for www18:
  • We need to get the estimator back to working. Set up vpn accounts to run the jobs?
  • How to decide which jobs go to the cluster and which jobs run in 18?
  • Run everything on the cluster by default (?) by setting the form to default to cluster. If the local machine is greyed out, but there is not default shown, what happens?
  • Stress test with the different options to decide what to do.
  • Emails to Phong and Mihael with suggestions on how to monitor these.
  • Find out if there is any way to know what node an analysis is running on so we could look at the activity on that node.
  • What are the other jobs that seem to be running.? Bob can kill a big job, but does not know what these other jobs are.
• What needs to be done if the system goes down under stress.
  • Scratch space that needs to be cleaned up (?). Email Phong about this again.
• Move posters autosave logic to production?
  • Add stars to poster file name and title.
  • Enter submits the poster?
  • Check for popup opening. Shouldn’t open popup with autosave.
  • Every 5 minutes.
• Drupal pending questions/issues (from telecon with Senem, Tom, Bob and Edit).
  • Phone call tomorrow at 11am.
• Code not currently working for job length estimation on large analyses.
  • Mihael sent Tom couple questions. Tom will answer them.

CMS

• CMS plots in the logbook.
  • Let Tom Mc Cauley know that changes may affect logbooks and posters.
  • Waiting on answer from Phong.

Cosmic

• Detector’s stacked flag when geometry changes.
  • Geometry ‘commit’ needs to change all the flags in the metadata.
  • Write some script to correct data that exists.