ATLAS UK Cloud Support

Name: ATLAS UK Cloud Support
Start: 2019-11-14T10:00:00+00:00
End: 2019-11-14T11:00:00+00:00
Location: Vidyo

Thursday 14 Nov 2019, 10:00 → 11:00 Europe/London

Vidyo

Tim Adye (Science and Technology Facilities Council STFC (GB)), Stewart Martin-Haugh (Science and Technology Facilities Council STFC (GB))

- 10:00 → 10:10
  
  Outstanding tickets 10m
  
  Open ATLAS UK GGUS tickets
  
  UK Cloud jobs over last week
  
  https://ggus.eu/?mode=ticket_info&ticket_id=144050 Lancaster deletion errors: no smoking gun, seems to have resolved. Will leave open for monitoring
  https://ggus.eu/?mode=ticket_info&ticket_id=144034 Birmingham Frontier issue: seems to just be a single user on T3, currently OK
  https://ggus.eu/?mode=ticket_info&ticket_id=143901 Sheffield failovers to Lyon: need more details - monster.shef.ac.uk is a frontend for VMs.
- 10:10 → 10:20
  
  Other new issues 10m
- 10:20 → 10:40
  Ongoing issues 20m
  - Centos 7 migration 5m
    
    Sussex: Currently working on implementing lcmaps for local users without Argus
  - Singularity 5m
  - Storageless sites 20m
    
    Sheffield
    
    Debate about direct I/O vs staging for Echo access from RAL - needs measurement. Will try to open RAL Echo monitoring outside firewall.
- 10:40 → 10:50
  
  News round-table 10m
  
  Dan: should get more storage soon, commission as new service - rsync from old cluster. Will mean temporary reduction in size, then increase.
  Elena: NTR
  Gareth/Sam: Ceph pool moved to new datacentre. Problems with GridFTP: machine not picking up full hostname for external interface. Also problems with certificates and xrootd/GridFTP. For new farm in data centre, may need new panda site or just new panda queues.
  Matt: NTR
  Patrick: NTR
  Stewart: Investigating forked jobs in Docker containers not having the correct P
  GID - Gareth suggested investigating container privileges and namespaces.
  Tim: NTR
- 10:50 → 11:00
  
  AOB 10m