Core Minutes 7/26/2011

• ScienceTools: (Richard) Jim is out of town so no report today.

• Pass7 Readiness: (Eric W) FSSC is on schedule for accepting Pass7 data next Monday, August 1. Already have all Pass7 data up to last Thursday.

  (Tom G.) will do another backfill today if nadir data problem is resolved quickly (which Warren thinks it likely will be) and a final one on Monday.

  (Tom G.) A skimmer problem is preventing 3 runs from being processed. Heather may look into it [but meanwhile Tom managed to work around it by retrying the runs several times until they went through].

• Documentation: (Richard, Heather) For now, Heather, Joanne and Jim have access to the Workbook site, which is on the Windows V disk. Joanne (and probably Jim) would be happier if it were moved to somewhere more readily accessible from unix; it's not clear whether this is feasible with the current server. (Richard) suggests we try updating something in the site to see if it works as expected.

• Pass8: (Tracy) A new GR tag is almost ready to go. Among other things, it features a return to our multiple scattering code, but with the newer version of G4, and implementation of multiple overlays on a single event.

• Event displays: (Leon) Dima is gone for a while, but just before he left he found and fixed a problem which was causing the display to be very slow. After the fix, performance was similar to Fred. However, there was an undesirable side-effect so he had to back it out. When he returns and has a chance to examine the situation more fully he should be able to reinstate it.

  (Tracy) Elliot & Co. are using yet another event display developed by Luca Baldini, bringing the total up to 4. This is the wrong direction.

  (Leon) will put together some instructions for installing WIRED, consulting Tony if need be.

• Hardware: (Richard) believes all new servers have arrived. One has already been installed for us. The next couple will be for EXO, then more will be installed for us (and for the groups who have been letting us borrow).

  We expect to make a tape order soon; waiting for other (e.g. EXO, KIPAC) to add on their requests. We'll order 1500 1 Tbyte tapes, of which about 1000 will be dedicated to Pass7 reprocessing.

• Core talk:

  • Windows SSH woes Despite Tom's new script that checks every 2 hours to see that the known_hosts files still exists and recreates if needed (which, he has confirmed, works as designeed), the file continues to disappear, even multiple times within a day. Sometimes this occurs in the process of a checkout, making a whole build fail - other times, the file has already disappeared and a build fails to check out any code. Similar problems for other required files were addressed by making the file read-only, but that's not possible for known_hosts since it may need to be modified. As we are at our wit's end, we are planning to request the unix-admins to modify the HOME environment variable for the glast user on windows to point to a local disk rather than the network Z drive. HOME determines where we store the .ssh directory. It is unknown if this will fix the issue, but we're willing to give it a try.

  • Another skimmer problem (Tom G.) is noted in JIRA SK-41.

  • Pass7 reprocessing support (Heather) At Tom G's request, Heather is working on updating the GR v17r35p23 branch to build with SCons. Fortunately this is only needed for Linux, simplifying the job considerably. Joanne has updated the tagCollector to work along branches. Meanwhile, Heather has about 8 more packages to update along the branch to capture all necessary SCons updates.

  • Externals path (Heather, Joanne) Since there were no objections we are proceeding with the plan to simplify externals paths by taking out the compiler component. Joanne wrote a script to automate the creation of new externals which Heather is about to try out [(Joanne) and did, immediately coming up against a flaw due to unrealistically simplistic assumptions about file hierarchy within externals. Should not be difficult to fix.]

  • Python 2.7.1 (Heather) There were some problems using it on rhel5 64-bit, now fixed. HEAD builds are now using this version.

  • MySQL upgrade (Heather) The final steps were put off until July, which it now is. We think our side is ready. She will contact Arash.

• Announcements

  • Accounts to vanish We plan to eliminate the Pipeline 1 prod and dev accounts glastdpf and glastdpt. If you know of a reason not to do this, speak up!

  • SCons meeting Tomorrow (Wednesday), 10:30 AM PDT

previous

minutes index

next

J. Bogart, Last Modified: