Core Minutes 8/2/2005  

Present:    Joanne Bogart, Toby Burnett, Jim Chiang Richard Dubois, Zach Fewtrell, Riccardo Giannitrapani, Tom Glanzman, Navid Golpayegani, Michael Kuss, Matt Langston, Julie McEnery, Chuck Patterson, Igor Pavlin, James Peachey, Leon Rochester, Tracy Usher

• DC2 code readiness: (Tracy) Made all-gamma and background runs at the end of last week and over the weekend. There were some crashes in CalValsTools; however, runs with the debugger on (so that the crash could be tracked down) didn't crash.

  Compared with previous, output for all-gamma shows some problems with CAL axis determination. Zach, at SLAC this week, will be investigating.

  Changes to CAL likelihood code caused a significant slow-down for these runs; since then it has been fixed. CAL moments analysis may need some work. (Richard) Raised noise level and changed threshold from 1.3 to 2.0. Tracy saw what should have been a 0.1 Mev hit show up as 40 Mev.  [Late-breaking news:  Zach has found and fixed the problem; Tracy has verified it's ok.]

  Toby has generated 5 million gammas and some background events at Seattle with a HEAD build. Although the gammas have the above problems, they are adequate for Toby's purposes.

  (Richard) Pruner, peeler, data server all seem fine. Pruning of background runs was taking about an hour [not really an hour, but an hour in the unit SLAC uses, tied to execution speed on no-longer-used hardware], at the limit of the queue, so Igor has raised the limit.

• Onboard filter: (Julie) Needs work. (Richard) Let's see where the dust settles after possible problems in CalXtalResponse are investigated. (Julie) At the very least, need a better value for pedestals that what is currently there. (Richard) Long-term, it looks like we'll be handling this by dynamically loading in one or another Flight Software shareable which will have the values represented in code.
• Pipeline migration: (Dan) plans to proceed as outlined below today, while I & T isn't running. He'll still have tomorrow morning if he needs it since data-taking isn't expected to start before tomorrow afternoon.
  1. A once over of tables to ensure that all fields are the same between PROD and TEST
  2. A diff of the stored procedure modules to see what new procedures need to be spliced in.
  3. Migration of a couple features from PROD (that aren't in test) to TEST
  4. Backup of existing PROD code.
  5. Tagging of TEST code in CVS.
  6. Fresh checkout of new tag into PROD area.
  7. Testing of code in PROD, including the new utilities he's written.
  8. Ensuring there's a hot spare crontab set up on one of the glast01 or glast03 boxes
  9. A few new utilities like delete task, etc. need to be written and tried in test.
• RM tour: (Navid) He described features of several types of pages, including main overview, summary for a tracked checkout package, such as GlastRelease, summary page for a particular build, such as GlastRelease v6r8 (tag rh9_gcc32), and red-dot summary page. Most of them are self-explanatory. Some of the newer features include
  • Ability to hide a particular build (if you have the privilege to do this and have logged in)
  • Diffs between releases. Defaults to diff against next-oldest, but at the bottom of the page there is a way to select something else.
  • Tag collector functionality is moved to red-dot summary pages. Log in for this as well. (Toby) Could we have an "are you sure?" warning before the operation completes, as was the case for old tag collector? (Navid) He'll see about putting it in.
  • Ability to delete a build, or parts of it (if you have the privilege to do this and have logged in). (Toby) Is it possible to abort a build in progress? (Navid) The complete delete will do this, albeit somewhat ungracefully. He'll look into neatening it up.
  (Richard) Navid will be presenting his work on RM at ADASS this fall.
• FRED: There is a request for some feedback after a source is selected, so you know the request went through. (Riccardo) will find a way to do this. He is near the end of a round of bug fixes, etc. When this is over, he'll make a new release which should become the standard one used by Gleam.

previous

minutes index

next

J. Bogart, Last Modified: 01-Jun-2010 15:45:01 -0700