Tier1 Operations Report 2017-11-08

From GridPP Wiki
Jump to: navigation, search

RAL Tier1 Operations Report for 8th November 2017

Review of Issues during the week 1st to 8th November 2017.
  • At last week's meeting we reported that CMS were switched over to using xrootd.echo.stfc.ac.uk meaning that CMS jobs used xrootd.echo.stfc.ac.uk as the primary means of accessing local data via xrootd. That change was reverted - but has since been re-applied. There was a further issue where long-lived CMS jobs needed to complete before the changes were picked up by all CMS batch jobs.
Current operational status and issues
  • There is a problem on the site firewall which is causing problems for some specific data flows. Discussions have been taking place with the vendor. This is having an effect on our data that flows through the firewall (such as to/from worker nodes).
  • CASTOR CMS Stager down. Investigation implies CPU error on physical box – “E1442 CPU2 machine check error. Power cycle AC.”)
Resolved Disk Server Issues
  • None
Ongoing Disk Server Issues
  • None
Limits on concurrent batch system jobs.
  • CMS Multicore 550
Notable Changes made since the last meeting.
  • Re-distribution of data in Echo onto the 2015 capacity hardware is ongoing. This is expected to complete in a few weeks.
  • CMS batch switched to use Echo as the primary means of accessing local data via xrootd. If data not found there it will fail over to Castor.
  • The Echo Gateways have had a parameter change that means the GridFTP gateways make better use of memory. This will enable the number of connections to each gateway server to be increased.
  • A start has been made on updating Castor tape servers to SL7.
Entries in GOC DB starting since the last report.
Service Scheduled? Outage/At Risk Start End Duration Reason
srm-cms.gridpp.rl.ac.uk, UNSCHEDULED OUTAGE 07/11/2017 06:00 07/11/2017 16:00 10 hours Hardware problems with Stager server - CMS Castor instance
Declared in the GOC DB
  • None
Advanced warning for other interventions
The following items are being discussed and are still to be formally scheduled and announced.

Ongoing or Pending - but not yet formally announced:

  • Re-distribute the data in Echo onto the 2015 capacity hardware. (Ongoing)
  • Update the LHCb Castor SRMs so as to be able to configure timeouts.

Listing by category:

  • Castor:
    • Update systems (initially tape servers) to use SL7 and configured by Quattor/Aquilon.
    • Move to generic Castor headnodes.
  • Echo:
    • Re-distribute the data in Echo onto the remaining 2015 capacity hardware.
    • Update to next CEPH version ("Luminous").
  • Networking
    • Extend the number of services on the production network with IPv6 dual stack. (Done for Perfsonar, all squids and the CVMFS Stratum-1 servers).
  • Services
    • The Production and "Test" (Atlas) FTS3 services will be merged and will make use of a resilient distributed database.

Open GGUS Tickets (Snapshot during morning of meeting)
GGUS ID Level Urgency State Creation Last Update VO Subject
131688 Green Urgent Assigned 2017-11-07 2017-11-08 CMS T1_UK_RAL has SAM3 CE critical > 2hours
131299 Green Urgent Assigned 2017-10-24 2017-10-24 CMS T1_UK_RAL HammerCloud failures
131213 Green Urgent In Progress 2017-10-19 2017-11-06 CMS Issues with fallback requests
130949 Amber Urgent Waiting for Reply 2017-10-06 2017-10-25 CMS Transfers failing to T1_UK_RAL_Disk
130207 Red Urgent In Progress 2017-08-24 2017-10-25 MICE Timeouts when copyiing MICE reco data to CASTOR
127597 Red Urgent On Hold 2017-04-07 2017-10-05 OPS Check networking and xrootd RAL-CERN performance
124876 Red Less Urgent On Hold 2016-11-07 2017-01-01 CMS [Rod Dashboard] Issue detected : hr.srce.GridFTP-Transfer-ops@gridftp.echo.stfc.ac.uk
124876 Red Less Urgent On Hold 2015-11-18 2017-11-06 NONE CASTOR at RAL not publishing GLUE 2
Availability Report
Day OPS Alice Atlas CMS LHCb Atlas Echo Comment
01/11/17 100 100 100 0 100 100 SRM Test failures.
02/10/17 100 100 100 0 100 100 SRM Test failures.
03/10/17 100 100 100 0 100 100 SRM Test failures.
04/10/17 100 100 100 0 100 100 SRM Test failures.
05/10/17 100 100 100 0 100 100 SRM Test failures.
06/10/17 100 100 100 0 100 100 SRM Test failures.
07/10/17 100 100 100 0 100 100 CMS: CE test failures caused by xroot data access problems to echo.
Hammercloud Test Report

Key: Atlas HC = Atlas HammerCloud (Queue ANALY_RAL_SL6, Template 845); Atlas HC Echo = Atlas Echo (Template 841);CMS HC = CMS HammerCloud

Day Atlas HC Atlas HC Echo CMS HC Comment
01/11/17 100 94 100
02/11/17 96 100 100
03/11/17 100 98 100
04/11/17 100 100 100
05/11/17 100 100 100
06/11/17 100 100 99
07/11/17 100 100 99
Notes from Meeting.
  • None yet