AF meeting 20th January 2011
Clemencic, Oliver Keeble, Sebastien Binet [remote], Marco
Cattaneo, John Harvey, Gabriele Cosmo, Predrag Buncic, David
Rousseau [remote], Rene Brun, Liz Sexton-Kennedy [remote],
Vincenzo Innocente, Benedikt Hegner, Fabrizio Furano, Fons
Rademakers, Andrea Valassi.
Line Management Issues (John)
Report from the LIM meeting (Benedikt)
News from projects
1 31/1/2011 @ 22:00)
- Fabrizio Furano (already known to many of us) will replace
Oliver Keeble at the AF. Fabrizio is now working in IT-GT-DMS (the
for DPM/LFC/gfal/lcg_util & FTS). He will following up the
deployment and releases in the AFS area of the Grid client libraries
required by the experiment's applications.
Management Issues (John)
- A new Technical Student, Anton Pytel, has started to work
Genser under the supervision of Witek Pokorski.
- NA61 has made a proposal to upgrade their old FORTRAN based
software. The SFT group will be
reviewing the proposal in 10 days or so. The review will be open and
the agenda will be circulated in due time.
what python modules exist in the external area (Benedikt)
of the status of the '_cov' platform in
the nightlies (Benedikt)
Planning Preparation (Pere)
- There was a call to the experiments to provide the 5 to 10 most
important requests/requirements for each of the projects (Simulation,
ROOT, SPI). This request for input has also been extended to include
the R&D projects (CernVM, Multi-core). We expect to receive the
initial input from the experiments in time for the next AF meeting. ==> Action ALL
presentation can be found
attached in the agenda page. Concerning the project management tool
JIRA, Sebastien asked whether there was an intention to replace
Savannah with it. The answer is that it is still too early to know if
this would be a good replacement (we need much more
experience) and full understanding of impact for the end-users of
Report from the LIM meeting (Benedikt)
Further information in the LIM
- LHCb reported the deprecation of scientific linux 4 and win-vc71
platforms in their software. Therefore the platforms will be removed
from upcoming LCG releases. Instead LHCb asked for 32 bit icc builds.
The slc4 build machines will be kept for the time being.
- ATLAS requested an LCGCMT_59b release for this week, to take
advantage of recent CORAL fixes. Boost version was reverted from 1.44
to 1.42 to be synchronized with the ATLAS Online/Offline.
- LHCb requested an LCGCMT_60 release to move to ROOT 5.28.00
News from Projects
- CernVM version 2.2 was released before Christmas. There has
been a lot of feedback during
holidays. In particular, performance issues reported by LHCb users
due to the new CernVM-FS catalog
organization. CMT is also not helping in there because wildcard
versions force to load too many catalogs. New version
(2.2.1) should include various improvements in this area.
- Despite the various power cuts during the CERN closure, they
manage to maintain the
critical services running. The repositories were mirrored to server
running on the Computer Center (read-only mode). Currently discussing
with RAL and BNL on repository replication
- There is an initiative to distribute the CernVM-FS client part
as part of the Scientific Linux.
- Preparing the Geant4 development release with bug fixes.
- Geant4 has migrated to SVN successfully. New SVN client version
is going to be available in the new version of SL5.
- Planning to use LCG_59b during the data-taking period this
- The new version of CORAL (patches) was made in December but it
has not been made available as part of any configuration. ATLAS is
make a release of this version available before a complete
is released in order to speedup the process. Marco C. argued that there
is a slot in the nightlies exactly for the purpose of providing patches
to the experiments. Sebastien replied that this slot may include many
other changes that they are not interested. ATLAS was encouraged
to use the slot in the nightlies with the patches, however under
special circumstances they may request a partial release of one
particular project. Sebastien will write down the actual request.
- The MacOSX and Windows on the nightlies need a fix ASAP. ==> action Benedikt
- ATLAS is currently building the release based on LCG_59b. From
next week they will start looking at LCG_60.
- They are just now starting the final validations of the release
they will use to start datataking. Barring last minute problems this
will have Geant4.9.4, ROOT 5.27/06 and (hopefully) they will be
fully (and only) 64bit in both their offline and online.
- As agreed they did start to do a bit of testing with ROOT
5.28/00, so far so good, and will continue this after the current rush
- Started using the new Geant4 in production during the holidays.
ROOT still having problems in their side with TMVA workaround.
- Preparing the build of the full software stuck
based on LCG_60.
They are going to report any problem as soon as they see any.
- The Windows builds are not in AFS. They have been
unable to test
the nightlies since December. The reason reported by Benedikt is that
the automatic job doing the copy crashes
with an authentication problem.
- Building their software based on ROOT 5.27.
- They has completed the migration to the CMake build