Agenda and Minutes, 19 September 2007
Agenda: Wed, 19 September 2007
Minutes
Attendance: Patrick Igor Greg Ben Duncan Felippo Adam Jolien Scott Warren Stuart Charles Kent Britta John Keith Z Carsten John W Kipp Ed Announcements: Carsten and John said the current time is better. Patrick asked if the intercall would be better. John has some problems with connecting to intercall. Stuart would prefer to have intercall work. Scott announced a workshop on LIGO workflows from the software infrastructure side to talk about the future and bring together everybody who is interested from LIGO on the first day. The Condor team will come along on the second day; Ewa Deelman should be able to make it too. Benoit has agreed to help fix compression bug with Ed for S6. Glue: Igor has been working on the SegFind server. The test server is currently running at LLO on a non-standard port. He has coalesced the science segments to test on that. Next step would be to add data qaulity segment information. For each DQ segment, create the next version which is like version 99 in segPageDump. Let segfind pick up highest coalesced version. Patrick asked Igor to send e-mail about what problems you are going to solve and when it will be done. Duncan asked about finding the flags that are on during a given time segment. Nothing else related to glue. DMT: After much ado and a random walk through rpm builds, everything is set up for rpm builds for dol including having everything set up correctly for both architectures. John sent Adam a message to that effect. Adam reported that he was able to do the build and he has published to the testing repo. He will test them right after the call. LAL/LALApps: nothing to report. Stuart asked about the report that there are bad frame files being produced by LAL. Jolien explained that he has not had time to look into it. John asked about the build instructions and getting them up to date. Jolien said he will try. Jolien also asked about switching to gsl-1.10 instead of 1.9. Jolien asked whether or not anybody knows if the GSL version 3 is on this new version of 1.10. No action was proposed. LDAS: Ed is continuing to write the system tests. LDG Client/Server: no progress report; Greg is waiting for VDT version 1.8.1 to be released. LDR: Scott fised a file descriptor leak that Murali found; also another bug noted by Duncan. LIGOTools: no report MatApps: Keith is looking at R2007b release. There are no big improvements in it. It supports big memory, no real improvements of impact to us. Metaio: Duncan reported that Alex has noted a possible problem. Stuart asked about the LIGOTools release: Duncan reported that this is in place. Patrick asked about moving Metaio to use a better xml reading tool. Duncan said it might be nice if we had people. Kipp reported that he thinks it could be done relatively quickly. What Kipp did is already in lsware and is pretty much working. Onasys: no report OSG-LIGO: Britta is just finishing up some old testing and testing on ITB. Kent asked if there are any open issues with new code. Duncan reported that there may be a bug in pipeline.py writing out the DAX. Britta is following up. AOB: Warren asked to get a report on the NR data format on a future agenda. Patrick will follow up on this. LSC Comp Comm has decided on the next OS with primary being CentOS 5 and debian etch. Discussion of Debian Etch support. Carsten and Patrick mentioned the ideas of the "Hackathon" and asked for feedback. Kent explained that they do it for OSG testbed release. Duncan reported on an experiment that he is working on to do fast data transfer between CIT and Syracuse. Kipp noticed that there is a HPC conference by a company to cell based clusters.
$Id: 070919.html,v 1.1 2007/09/26 18:03:15 patrick Exp $