NAF User Committee Monthly Meeting

Europe/Berlin
Yves Kemp (IT (IT Systems))

IT: Christoph, Stefan, Yves

CMS: Johannes, Hartmut

Belle: Andreas

Luxe: Federico

Atlas: Stephen

 

IT Status

1) Jupyter: 

  • More logs? Difficult, since there is not much information available

2) Memory problems:

3) Behaviour change

  • Experiment support should be involved (latest in the "escalation" emails, maybe also in the automated emai)

4) WGS

  • test jobs: maybe adding DUST ls check (take offline)
  • check WGS hardware / sizing

5) DUST

  • Ganesha issues not observed in other environments

6) DUST migration

  • Plan: Luxe migrate with ATLAS, ILC and AXION
  • OK with the migration plan
  • Symbolic links:
    • ATLAS for initial migration period
    • Experiment admins to create the symlinks
    • Duration: maybe until beginning of May'25
  • BeeGFS group volumes

7) Miscellaneous 

  • Sectigo situation
  • NAF partial shutdown over change of year

 

CMS report

1) Rethink Ticket system

2) Survey on NAF usage

  • more memory intensive applications
  • columnar analysis
  • Production of custom NanoAOD are using CMS grid, not the NAF.

3) Ideas:

  • monthly NUC meetings again
  • Feedback
  • Change in usage pattern
    • e.g. less production like slimming/slimming of official data sets

 

General discussion

  • higher frequency of emails to NAF users: welcome
  • next NUC: 28.1.25 @13:00
  •  
There are minutes attached to this event. Show them.
    • 13:00 13:40
      NAF status 40m
      Speaker: Yves Kemp (IT (IT Systems))
    • 13:40 14:00
      CMS slides 20m
      Speaker: Hartmut Stadie (UNI/EXP (Uni Hamburg, Institut fur Experimentalphysik))