Login

ATLAS UK Cloud Support

→ Europe/London
Vidyo

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

● Outstanding tickets

GGUS #142774  Matt reported that Lancaster have a sick disk server. It is being drained, but this is a slow process. The data is safe and access is OK as long as there is not too much load.

● Birmingham/Cambridge XCache

John: XCache sites received a request from Ilija Vukotic (ATLAS) to set up monitoring. Cambridge and one other site have done it so far. John forwarded the email thread to atlas-support-cloud-UK@cern.ch.

● Diskless sites

All Cambridge and Sussex disks (apart from UKI-SOUTHGRID-CAM-HEP-XCACHE, which we keep) have finished draining and have now been removed from Rucio. As far as ATLAS is concerned, those disks can be decommissioned.

John noted that there were a few files left on the Cambridge SCRATCHDISK. These can be removed.

● Centos 7 migration

RHUL: CE queues set up but not yet working. No PanDA queues yet. Elena will ask Alessandra.
Sussex: Now storageless, so can move forward with CentOS7 migration. Patrick said that Sussex are likely to overrun the deadline by a bit.
Glasgow: Sam reported that Gareth and Alessandra have talked. CentOS7 queues are setup, but still need resources.
Elena confirmed that all other sites are OK.

● Singularity

Edinburgh, Durham, and Sussex need to install Singularity. Can install locally or use CVMFS. Durham and Sussex will probably use CVMFS.

● News round-table

Dan: NTR
Elena: NTR
John: SE will go into downtime next week. When decommissioned will shut down BDII.
Matt: is on leave for next 2 weeks. Peter is around.
Patrick: NTR
Peter: is around, but away Thurs/Fri next week.
Sam: migration proceeding at Glasgow
Tim: Added next (3/4) 0.5PB allocation to RAL DATADISK. Wants to understand REBUS CPU capacity numbers, eg. why some sites like Manchester have ATLAS capacity = 0.

There are minutes attached to this event. Show them.
    • 10:00 → 10:10
      Outstanding tickets 10m
      Open ATLAS UK GGUS tickets

      GGUS #142774  Matt reported that Lancaster have a sick disk server. It is being drained, but this is a slow process. The data is safe and access is OK as long as there is not too much load.

    • 10:10 → 10:30
      Ongoing issues 20m
      • Birmingham/Cambridge XCache 5m
        Cambridge XCache JIRA ticket

        John: XCache sites received a request from Ilija Vukotic (ATLAS) to set up monitoring. Cambridge and one other site have done it so far. John forwarded the email thread to atlas-support-cloud-UK@cern.ch.

      • Diskless sites 5m
        Sussex disklessness JIRA ticket

        All Cambridge and Sussex disks (apart from UKI-SOUTHGRID-CAM-HEP-XCACHE, which we keep) have finished draining and have now been removed from Rucio. As far as ATLAS is concerned, those disks can be decommissioned.

        John noted that there were a few files left on the Cambridge SCRATCHDISK. These can be removed.

      • Centos 7 migration 5m
        Centos 7 deployment Twiki

        RHUL: CE queues set up but not yet working. No PanDA queues yet. Elena will ask Alessandra.
        Sussex: Now storageless, so can move forward with CentOS7 migration. Patrick said that Sussex are likely to overrun the deadline by a bit.
        Glasgow: Sam reported that Gareth and Alessandra have talked. CentOS7 queues are setup, but still need resources.
        Elena confirmed that all other sites are OK.

      • Singularity 5m
        Singularity status (Edinburgh/Durham

        Edinburgh, Durham, and Sussex need to install Singularity. Can install locally or use CVMFS. Durham and Sussex will probably use CVMFS.

    • 10:30 → 10:50
      News round-table 20m

      Dan: NTR
      Elena: NTR
      John: SE will go into downtime next week. When decommissioned will shut down BDII.
      Matt: is on leave for next 2 weeks. Peter is around.
      Patrick: NTR
      Peter: is around, but away Thurs/Fri next week.
      Sam: migration proceeding at Glasgow
      Tim: Added next (3/4) 0.5PB allocation to RAL DATADISK. Wants to understand REBUS CPU capacity numbers, eg. why some sites like Manchester have ATLAS capacity = 0.

    • 10:50 → 11:00
      AOB 10m