Core Minutes 28 October 2014

Core Minutes 10/28/2014

Present: Joe Asercion, Joanne Bogart, Jim Chiang, Richard Dubois, Warren Focke, Tom Glanzman, Heather Kelly, David Landriu, Leon Rochester,

ScienceTools: (Jim) No news.
FSSC: (Joe) No news.
Reprocessing: (Tom G.) Fermi Pass8 reprocessing has been updated through September 2014. This was a 5-month update.
Hardware: (Tom) Yemi has been putting together a global scratch system based on the Lustre high-performance file system. It had been visible only to the bullet cluster (via high-performance infiniBand networking) but is now visible also to the 'dole' batch nodes and the rhel6-64 interactive login machines (via a slower NFS-like mount). Soon, that visibility should be extended to all batch machines. Fermi currently has a 50 TB scratch space which is accessible to pipeline jobs (or users in the YP pipeline group). This scratch is managed automatically with a 7-day retention period.
Pass8: (Leon) No news at present. He expects to get started on the Pass7/Pass8 reconstruction thing (that is, generate MC events with p8, then compare recons of those events from p7 and p8) and have something to report next time.
Core talk:
- Workbook (Heather) will update to point to FSSC documentation where appropriate.
- Mac (Joanne) There has been no progress since last time on the VM. Meanwhile, another possible user for the machine has surfaced. Andy Hanushevsky would like to be able to use it for builds for LSST now that SCS has decommissioned all their Macs. He needs cmake and gdb (not originally installed on our machine), some disk space, and an account. All of these things are possible without much effort — in fact, since we might also have use for cmake and gdb I went ahead and installed them, using Macports at Tom Stephens' suggesion — but I'm worried about mission creep. Andy has asked LSST if they can make a machine available; he's waiting for their response.
- New flight software build (Heather) The new FSW external (rhel5 32-bit only) has been unpacked and installed in the proper place. Everything should be ready for Jim to make a new build of CHS.
  
  It is still an open question whether we need a new obf external or not on Linux (it is definitely not worth the effort on Windows). [(Joanne) After the meeting Heather pointed me to the Flight Software release notes, which enumerate packages that have been changed. None of them are in our obf external so there is no reason for us to make a new version.]
- rhel6 validation & obf crash (Joanne) spoke to JJ. He was not available at the time for a joint debugging session, but might be in the future. It could be useful to look around in a core dump on my own, first. The crash is in Flight code, but it could easily be because our stuff is passing bad arguments. (Heather) knows what job options Johan used so could probably produce one.

minutes index

J. Bogart, Last Modified: