ATLAS UK Cloud Support

Thursday 15 Apr 2021, 10:00 → 11:00 Europe/London

Zoom

Tim Adye (Science and Technology Facilities Council STFC (GB)), James William Walder (Science and Technology Facilities Council STFC (GB))

https://cern.zoom.us/j/98434450232

Password protected (same as (new) OPs Mtg)

● Outstanding tickets

• Outstanding tickets

  • 151141 TEAM atlas UKI-NORTHGRID-LANCS-HEP less urgent NGI_UK in progress 2021-04-15 08:25:00 UKI-NORTHGRID-LANCS-HEP deletion errors WLCG

    • Current set of files declared as lost and also set to be deleted from the DPM namespace

  • 151098 USER atlas RAL-LCG2 urgent NGI_UK in progress 2021-04-12 09:20:00 High failure rate at RAL-LCG2_TEST WLCG

    • Awaiting deployment to full cluster; can switch to multi-job pilots once done

---

● CPU

• • RAL

    • HC test, together with CMS releasing caps results in dropped cores
    • CoreCount still not correct; need to follow up in Cric

  • Northgrid

    • Slight variation for Lancs; due to other users, and no issue

  • London

    • RHUL was in Downtime (Resumed this morning; but no jobs yet); Hammercloud issues?

  • SouthGrid

    • SUSX starting to run jobs

  • Scotgrid

    • No issues; ECDF with occasional HC failures

• Other new issues / tasks

---

● Ongoing Items

• CentOS7 - Sussex

  • Current issues appear to related to location and size of disk space for the jobs to run in.
    • Change of configuration to the starting dir made today
  • Space per core requirements documented here:
    • https://twiki.cern.ch/twiki/bin/view/AtlasComputing/SitesSetupAndConfiguration#Worker_Node_hardware_resources
  • Patrick noted strange issues wth various accounts
  • node219 seems more problematic than other nodes
  • To continue follow-up as needed as necessary

• TPC with http

  • Work continues at RAL with Xrootd

• Storageless Site test / storage decomissioning (Oxford)

  • Automated part of process has begun
  • Issue with (atlas) site in cric with different write-back points; Aim to transition main production across to RAL endpoint asap
  • LOCALGROUPDISK will more to QMUL
  • JW to follow-up on necessary items

• Glasgow DPM Decommissioning

  • To see if any further progress done or needed.

• ATLAS: Site Availability/Reliability reports: Glasgow

  • New snow ticket updated; cern team say work will be done in next ‘few weeks’

 

 

---

● News round-table

• Dan
  • NTR
• JW
  • NTR
• Matt
  • New kit is racked; and needs plumbing in
• Patrick
  • NTR
• Peter
  • NTR
• Sam
  • NTR
• Vip
  • NTR

 

 

10:00 → 10:20 Status

10:20 → 10:40 Ongoing Items

• CentOS7 - Sussex

  • Current issues appear to related to location and size of disk space for the jobs to run in.
    • Change of configuration to the starting dir made today
  • Space per core requirements documented here:
    • https://twiki.cern.ch/twiki/bin/view/AtlasComputing/SitesSetupAndConfiguration#Worker_Node_hardware_resources
  • Patrick noted strange issues wth various accounts
  • node219 seems more problematic than other nodes
  • To continue follow-up as needed as necessary

• TPC with http

  • Work continues at RAL with Xrootd

• Storageless Site test / storage decomissioning (Oxford)

  • Automated part of process has begun
  • Issue with (atlas) site in cric with different write-back points; Aim to transition main production across to RAL endpoint asap
  • LOCALGROUPDISK will more to QMUL
  • JW to follow-up on necessary items

• Glasgow DPM Decommissioning

  • To see if any further progress done or needed.

• ATLAS: Site Availability/Reliability reports: Glasgow

  • New snow ticket updated; cern team say work will be done in next ‘few weeks’

 

 

10:40 → 10:50 News round-table

• Dan
  • NTR
• JW
  • NTR
• Matt
  • New kit is racked; and needs plumbing in
• Patrick
  • NTR
• Peter
  • NTR
• Sam
  • NTR
• Vip
  • NTR

 

 

10:50 → 11:00 AOB