Core Minutes 6/16/2015

• ScienceTools: (Jim) No news.

  See Science Tools Development Notes for current activity.

• FSSC: (Joe) Preparation for the ScienceTools release supporting P8 is ongoing. It's still on track for June 24th. A minor issue on newer Mac platforms (Mavericks, Yosemite) concerning plotting still needs to be resolved. Work is also in progress to support the Background Estimator.

• Reprocessing: (Tom) Another backfill, good through June 3rd, is complete, leaving one last 3-week backfill to bring us up to the 24th. (Richard) How long will this take? (Tom) If all goes perfectly smoothly, about four days: two for the main recon and one each for generating merit and FITS.

• Level 1 readiness: (Richard) Maria Elena encountered a problem processing LCI data, described in JIRA LPATE-188. Steve Tether discovered the cause (improper handling of certain variables on 64-bit machines) and has implemented a fix. (Warren) It can't be used to digitize LCI data until it's been incorporated into a new CHS release, but we won't be taking charge injection data for another 6 months. It will be ready by the time we need it.

• Hardware: (Tom) The new user disk arrangement is basically running smoothly, but there is still some question whether ganglia is doing an adequate job of monitoring behavior. We need complete and reliable reporting in order to detect and diagnose problems.

  Some xrootd servers have old flaky Seagate drives. Work is in progress to swap in better drives.

• News from C&A: (Leon) couldn't make the meeting but sent this report:

  Two interesting presentations at the C&A meeting: David Green on the proton spectra up to 6 TeV of unfolded energy, and Thibaut D. on the low-energy sky, where low means down to ~30 MeV. In the first, our results are compared to AMS, and in the second to COMPTEL!

• Core talk:

  • Doxygen Generation (Joanne) tracked down the problem with doxygen generation for GlastRelease (it was already ok for ScienceTools). The RM database configuration was such that the doxygen generation step (separate from and following the SCons build step) would only occur for rhel5 64-bit builds, but rhel5 64-bit is not a supported platform for GR. After changing the setting for GR to use rhel6 64-bit instead the doxygen generation step ran correctly when invoked manually. It is most likely fixed, but it would be nice to see it work automatically as part of RM. (Richard) Shouldn't we also change ST to use rhel6 64-bit? (Joanne) if there are no objections [there weren't] will do so as soon as correct operation for an RM buld of GR is confirmed.

    See details at JIRA GRINF-69.

previous

minutes index

next

J. Bogart, Last Modified: