Architects Forum Meeting

Thursday 8 Feb 2018, 14:00 → 16:00 Europe/Zurich

32/S-C22 (CERN)

Dial-in numbers: 71400 (CERN), +41227671400
Vidyo meeting Numeric Extension: 10287102

Present: Patricia Mendez Lorenzo, Charles Delort, Geri Ganis, Benedikt Hegner, Michael (IT), Alejandro (IT), Giulio Eulisse, Graeme Stewart, Jakob Blomer, Marco Cattaneo, Walter Lampl, Gerardo Ganis, Gabriele Cosmo, David Lange

Excused: Pere Mato, Patricia Mendez Lorenzo

Line Management Issues (Gerri)

• Two new people in the group - Enric started as staff

• Technical student in CernVM. Docker interface consolidation

• Plan of work of the SFT teams presented as of 1st of March.

• EP R&D working group chaired by Jakob and Graeme. Kick of took place

Report from the LIM Meeting

• Preparing the 93 release

Status of Projects

• CERNVM

  • CernVM workshop last week. Quite happy. Feedback from experiments. Immediate items - new CernVM image. Meltdown / spectre fixes cost a lot of time for testing CVMFS 2.5. Release in a month

  • Pleasant surprise that we know what users want and do.

    • Faster and easier publication would be good. Requires talking to IT. maybe SSD back

    • Giulio: was the point of smaller granularity in permissions mentioned? J: a few developments. User name space. Portals. S3 endpoint for a particular area. Pushing tarballs into the area. Porjh

• Simulation

  • Since last time working on new patch to 10.4. Preparing the work plan for 2018. Migrating to gitlab within two/three months.

  • Request from LHCb and FCC to have latest releases of Geant4 in LCG releases. Plus latest VecGeom. LHCb moving to 10.3.

  • LHCb comment: builds on very latest compilers would be useful

  • Giulio: why gitlab and not github? → Control system; granular control of access; public releaes mirrored to github anyhow

• Persistency

  • 3.2.0 as new CORAL/COOL. Qt4 to Qt5 migration.

• ROOT

  • Tag in preparation 6.12 frozen yesterday. Planning meeting on wednesday

 

Experiments’ Feedback

• ALICE

  • Finally moving to root 6 in production releases. Business as usual A few bugs discovered and fixed.

• LHCb

  • Nothing special. Split software in two main branches. 2018 data taking and beyond for analysis; latest/greatest for everything; trigger as focus now.;

  • Plenty of hackathons. Huge improvements in SW by looking closer at things.

• ATLAS

  • Two branches for future and data taking. Reprocessing for conditions update. Master branch working towards AthenaMT. LCG 92 recompile linker problem. New linker, but not shipped with lcg itself.

  • RPM containing binutils provided by Patricia. Various other problems.

• CMS

  • Nothing to report

  • Working on production releases

 

• IT

  • NTR

  • Meltdown advice

  • Question by marco: ITUM move to CentOS7 presented. Maybe something more modern by 2020?

• HSF

  • Published CWP in December; still gathering signatures

  • Workshop in Naples advert

  • Other news: packaging group is rather active in writing use cases. More news on that later.

14:00 → 14:10 Report from the LIM meeting

Speaker: Dr Patricia Mendez Lorenzo (CERN)

14:10 → 14:20 Status of Projects

14:20 → 14:30 Experiments Feedback