LHCb Core Software Meeting
Date and Location
2015-06-24
10:30 - 11:00
CERN (2-R-030)
Attendees
Colas Pomies,
Gloria Corti,
Illya Shapoval,
Joel Closier,
Marco Cattaneo,
Marco Clemencic (minutes)
Subjects
News
Round Table
Marco Cl.
- Jenkins / Nightly Builds
- moved git repository to gitlab (LBCORE-829)
- upgraded server and plugins
- re-organization of Jenkins' jobs
- added priority lane for release jobs
- An ATLAS developer will help testing Gaudi for ATLAS
- Issue with Beat's script to push alignments and calibrations to Offline CondDB fixed yesterday, but there is another issue to be fixed (affecting a very improbable corner case)
Illya: checked the CondDB deployment. Yesterday there was a burst of updates.
Marco Cl. Beat added several runs in one go, after the fix.
Marco Cl.: Joel, can we access
http://eoslhcb.cern.ch
form outside CERN? If so, we should move the CondDB deployment files to EOS (instead of WEBAFS).
Marco Ca.
- Issue with LoKi dictionaries still present (lhcb-lcg-head used LCG 77 instead of dev4)
- Marco Cl.: the issue was fixed yesterday, but I forgot that I changed lhcb-lcg-head. Will be fixed ASAP.
- Issue with a test in DetCond to be checked by Marco Cl.
- Strange problem with Muon calibration in Brunel tests
- Marco Cl.: there was a retroactive change in a (new) L0MUON condition which should not affect the
- Marco Ca.: is this change logged somewhere? NO. Can we know what changed in the new calibration conditions? In principle yes, but we need to introduce (automatic) tags to simplify the checks.
Gloria
- Problem with permission scheme used by LHCb Simulation: it seems that it was changed. Ben is following up.
Illya
- Increased the frequency of CondDB deployment cron. It cannot be reduced more than that (it currently takes ~7-8 minutes).
- Ben modified the deployment of CondDB to CVMFS to avoid empty transactions and increased the frequency too.
- Ariadne
- moving to gitlab.cern.ch
- introduce new type of node to track compatibility on Event Data Format
--
MarcoClemencic - 2015-06-24