Tier1 and Tier-2 Weekly Report 2017-09-05

Tier1 and Tier-2 Weekly Report 2017-09-05

General News

NO MEETING - ADC Technical Coordination Meeting
ADC Weekly Meeting [20170905]
- MC Production
  - Usual ~250k slots being used
  - MCProd currently 76% global share
  - Mixture of mc15, mc16a, mc16c
    - MC15: Evgen is ~50/50 mc15/mc16
    - MC16: Four mini-campaigns running
  - ~400M events pending/~75M waiting approval
  - Not much simul remaining
  - Pileup can be submitted when grid empties
  - Non-grid resources:
    - ORNL Titan mcore (backfill)
    - NERSC Cori mcore (dedicated)
    - BOINC mcore (private Tier-2) - low priority request shared with grid
- Group Derivation Production - no report
- Reprocessing - no report
- CRC/ADCoS Report - no report
- Software Baseline
  - Last week some sites had a problem with the HEP_OSlibs old version
  - Suggested we should have a software baseline page to avoid these problems
  - ATLAS doesn't have one, but WLCG does
    - https://wlcg-mw-readiness.cern.ch/baseline/current/
    - So far only middleware but can be extended upon request
      - HEP_OSlibs, singularity, etc
  - WLCG also presents sw/mw announcements at ops meetings
    - Next will be next week
    - Not all sites attend the WLCG ops meeting or read the minutes (but the work is already done we should avoid duplicate)
  - Repeat changes WLCG announcements to ATLAS specific meetings
  - Ask for software to be added if needs to be
  - Sites should bookmark the baseline page
  - Page link can be added to the VO ID card and SiteSetup twiki
    - ATLAS VO ID card https://operations-portal.egi.eu/vo/view/voname/atlas
    - ATLAS Sites Setup and Configuration twiki https://twiki.cern.ch/twiki/bin/view/AtlasComputing/SitesSetupAndConfiguration
- Pilot Update v70.4
  - Return proper error code to wrapper when out of local disk space (at pilot startup)
    - Previously led to lost heartbeat on ND
  - Changes for unified PanDA queues
    - Now identifying user jobs from trf name instead of queue name (ANALY_*)
    - Will later use new job identifier (user/prod), discussed with Tadashi
  - Benchmarking now switched on for user jobs
    - Previously only for production jobs due to jobReport requirement (not produced by most user jobs)
    - Now sending benchmark results to intermediary Elastic Search service (i.e. jobReport not used)
  - Setup fixes for R21 on ND
  - Logs to OS
    - Support for OS log transfers in mv mover
    - Fix for wrong naming convention
    - Logs registered for all jobs, i.e. now also for failed jobs
  - Now reporting metadata for zipped ES files
CA Cloud Squad Actions
- User DaTRI requests approved / processed (last week)
Open Issues and Ongoing Discussions
- Availability & Reliability Reports [20170328]
  - 2017 Reports - click down to find reports of format WLCG_All_Sites_ATLAS_MMMYYYY.(html|pdf)
- Sites should be providing the 2016 storage pledge.
  - REBUS Site Capacities
  - REBUS Federation Resource Pledges
- ATLAS Canada HammerCloud Tests
WLCG and General Computing News
- Security News
- gLite/EMI News
ROC_Canada News
- Current shift