ATLAS UK Cloud Support

Thursday 4 Mar 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

• 150820 UKI-LT2-RHUL less urgent in progress 2021-03-03 15:18:00 UKI-LT2-RHUL: 0% Transfer and deletion efficiencies
  • Power incident in center. Possible problems in bringing up some hardware.
• 150775 UKI-SCOTGRID-GLASGOW less urgent in progress 2021-03-04 04:45:00 UKI-SCOTGRID-GLASGOW transfer and deletion errors
  • External link going down to 1 Gb
  • Many routes of campus; might be cause of issue
  • Association to large numbers of small files attempted to be transfered in via FTS
  • Restart of gridFTP services ‘cures’ problem
• 149362 UKI-SOUTHGRID-RALPP urgent in progress 2021-02-18 20:00:00 ATLAS CE failures on UKI-SOUTHGRID-RALPP-heplnx207
  • No progress
• 146651 RAL-LCG2 urgent on hold 2021-02-16 17:37:00 singularity and user NS setup at RAL
  • No progress
• 142329 UKI-SOUTHGRID-SUSX top priority on hold 2021-01-20 20:29:00 CentOS7 migration UKI-SOUTHGRID-SUSX
  • Issue with arex, due to change of hypervisor
  • Networking should be ok now.
  • Potential issues for
  • A few (3/4) different hardware sets; Monitor each set as provisioned.
  • Patrick to update ticket once confirming network

---

● CPU

• General problem with non-aCT sites, due to Harvester issue on Sunday evening, casuing worldwide job reductions

• Weds. another dip in central production; presumed due to Condor update (for ARC / GDPR fixes)

• RAL

  • Job reductions due to:
    • Capped to 100% to enable more jobs for LHCb
    • ATLAS demands for more SCORE user jobs;

• Northgrid

  • Lancs: Fairshare (stable to O(3k)), may want to try and bounce other users.

• London

  • OK

• SouthGrid

  • OK

• Scotgrid

  • OK

• BHAM also not running for LHCb, No update.

• CAM to go into long downtime; long-term future to be discussed.

---

● Ongoing Items

• CentOS7 - Sussex

  • Making good progress now (see abouve )

• TPC with http

  • Moving to WebDav for wan / lan / TPC.
  • LHCb enabled everywhere, except for RAL, GLA, QMUL(?)
  • CMS: Similar RAL, GLA(~ no local storage), QMUL (use IC)
  • DPM / xrootd; different libCurl versions
    • Push for xrootD sites now
    • Discussion on tokens occured
    • Users still tend to complain on not liking the grid.
  • GLA DirectIO possibility if WAN xrootd enabled. Needs a Cache inplace.
    • LAN - just to test that LAN works
    • WAN to be main focus
  • QMUL - prefer to wait (should already work, but server not quite powerful enough) for new hardware.
    • gridFTP, one main server that redirects for the actual transfers
    • WebDav, might need some configuration to do the same as gridFTP. One powerful node should however work fine.

• Storageless Site test / storage decomissioning (Oxford)

  • Sam to get to Vip today updated config for Rules refinment.
  • ECDF ES xrootd cache monitoring up, but not seeing Xcache transfers.

• ECDF volatile storage

  • Rob reconfigured the site
  • JW to make the necessary ATLAS updates

• Glasgow DPM Decommissioning

  • LOCALGROUPDISK done; Datadisk residual data remain

• ATLAS: Site Availability/Reliability reports: Glasgow

  • Is this done?

---

● News round-table

• Vip

  • On leave next week

• Dan

  • NTR

• Matt

  • Disk server reboot - not come up;
    • Couple of TB data
    • 11k files might be declared lost

• Peter

  • (Needed to leave before end)

• Alessandra

  • NTR

• Sam

  • NTR

• Gareth

  • Will be leaving GridPP in April

• JW

  • NTR

• Patrick

  • NTR

• Rob

  • NTR

10:00 → 10:20 Status

10:20 → 10:40 Ongoing Items

• CentOS7 - Sussex

  • Making good progress now (see abouve )

• TPC with http

  • Moving to WebDav for wan / lan / TPC.
  • LHCb enabled everywhere, except for RAL, GLA, QMUL(?)
  • CMS: Similar RAL, GLA(~ no local storage), QMUL (use IC)
  • DPM / xrootd; different libCurl versions
    • Push for xrootD sites now
    • Discussion on tokens occured
    • Users still tend to complain on not liking the grid.
  • GLA DirectIO possibility if WAN xrootd enabled. Needs a Cache inplace.
    • LAN - just to test that LAN works
    • WAN to be main focus
  • QMUL - prefer to wait (should already work, but server not quite powerful enough) for new hardware.
    • gridFTP, one main server that redirects for the actual transfers
    • WebDav, might need some configuration to do the same as gridFTP. One powerful node should however work fine.

• Storageless Site test / storage decomissioning (Oxford)

  • Sam to get to Vip today updated config for Rules refinment.
  • ECDF ES xrootd cache monitoring up, but not seeing Xcache transfers.

• ECDF volatile storage

  • Rob reconfigured the site
  • JW to make the necessary ATLAS updates

• Glasgow DPM Decommissioning

  • LOCALGROUPDISK done; Datadisk residual data remain

• ATLAS: Site Availability/Reliability reports: Glasgow

  • Is this done?

10:40 → 10:50 News round-table

• Vip

  • On leave next week

• Dan

  • NTR

• Matt

  • Disk server reboot - not come up;
    • Couple of TB data
    • 11k files might be declared lost

• Peter

  • (Needed to leave before end)

• Alessandra

  • NTR

• Sam

  • NTR

• Gareth

  • Will be leaving GridPP in April

• JW

  • NTR

• Patrick

  • NTR

• Rob

  • NTR

10:50 → 11:00 AOB