ATLAS UK Cloud Support

Thursday 3 Feb 2022, 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

• 155856 TEAM atlas RAL-LCG2 less urgent NGI_UK in progress 2022-02-03 09:50:00 RAL-LCG2: deletion errors EGI

  • RAL disk is full, large numbers of deletions of small files.
  • DDM to try and launch larger file deletions.
  • Generally just very busy …

• 155430 TEAM atlas UKI-SCOTGRID-ECDF less urgent NGI_UK in progress 2022-01-13 14:27:00 UKI-SCOTGRID-ECDF transfer and deletion errors EGI

  • Awaiting large file list to arrive, do be declared as lost

• 155141 TEAM atlas UKI-LT2-Brunel less urgent NGI_UK in progress 2022-02-01 18:23:00 Transfers from UKI-LT2-Brunel fail with “Internal Server Error” EGI

  • Needing the ok from Brunel that they’ve cleaned the files from their namespace

• 154806 TEAM atlas UKI-LT2-QMUL less urgent NGI_UK in progress 2022-01-28 09:16:00 UKI-LT2-QMUL SOURCE transfer failures EGI

  • SSL errors remain large source of transfer failures
  • Restarting of some disk server needed
  • Storm added some dev version to repo, causing SRR issues, now reverted
  • Additional storage being added to the site

• 154543 TEAM atlas UKI-SCOTGRID-ECDF urgent NGI_UK in progress 2021-12-08 12:35:00 DPM storage ACL configuration EGI

  • Still needing an update

• 154436 TEAM atlas RAL-LCG2 very urgent NGI_UK in progress 2022-02-03 08:15:00 RAL Echo Davs developments EGI

  • More file transfer failures reported due to known multihop limitations

• 153367 TEAM atlas RAL-LCG2 urgent NGI_UK on hold 2021-12-01 15:37:00 HTTPS on RAL CTA EGI

  • Disucssion on TAPE and Disk tests, following the recommendations of data challenges.

---

● CPU

• • RAL

    • Reduced capacity for patching camapaign finished; ATLAS running well, so far …

  • Northgrid

    • NTR

  • London

    • QMUL returning after a few issues.
      • Dark data reported
      • Storm dev version was in repo, also doing the SRR?

  • SouthGrid

    • Xcache issues, followed by power glitch on weekend

  • Scotgrid

    • Xrootd service needed restarting. 503 rucio errors masked this from quickly being spotted

---

● Ongoing Items

• TPC with http

  • ATLAS wanting to drop gridFTP by March:
    • Castor <-> Echo would still be ok (if not moved to CTA by then)
    • Echo writes remain on gridFTP (predominantly);
      • With additonal hardware, should be able to switch once installed
    • Glasgow; gridFTP; move to cephfs, or, use patched davs version from RAL
      • Could consider to set up additional SE, and move data across managed by DDM
      • JW to liaise with DDM

• Storageless Site test (Oxford)

  • NTR (discussed in Storage Mtg)
    • JW to get avaialble plots together by end of Feb

• LANCS Storage migration

  • Viable endpoint for functional tests should be avaialble by today
  • JW to make final updates to CRIC and run checks before informing DDM

 

 

---

● News round-table

• Alessandra

  • NTR

• Dan

  • NTR (discuessed above )
  • Alessandra helping with GPU queue setting

• Gerard

  • NTR

• Matt

  • Discussed endpoint avaialbility

• Sam

  • NTR

• Stephen

  • NTR

• Vip

  • NTR

 

 

1 Status

2 Ongoing Items

• TPC with http

  • ATLAS wanting to drop gridFTP by March:
    • Castor <-> Echo would still be ok (if not moved to CTA by then)
    • Echo writes remain on gridFTP (predominantly);
      • With additonal hardware, should be able to switch once installed
    • Glasgow; gridFTP; move to cephfs, or, use patched davs version from RAL
      • Could consider to set up additional SE, and move data across managed by DDM
      • JW to liaise with DDM

• Storageless Site test (Oxford)

  • NTR (discussed in Storage Mtg)
    • JW to get avaialble plots together by end of Feb

• LANCS Storage migration

  • Viable endpoint for functional tests should be avaialble by today
  • JW to make final updates to CRIC and run checks before informing DDM

 

 

3 News round-table

• Alessandra

  • NTR

• Dan

  • NTR (discuessed above )
  • Alessandra helping with GPU queue setting

• Gerard

  • NTR

• Matt

  • Discussed endpoint avaialbility

• Sam

  • NTR

• Stephen

  • NTR

• Vip

  • NTR

 

 

4 AOB

ATLAS Monitoring (UK Cloud)

Next Week S&C

Storage ACLs