Skip to content

OSG Technology Area Meeting, 13 August 2018

Coordinates: Conference: 719-284-5267, PIN: 57363; https://www.uberconference.com/osgblin
Attending: BrianL, Carl, Derek, Edgar, Marian, Mat, Suchandra, TimT

Announcements

Triage Duty

  • This week: Edgar
  • Next week: BrianL
  • 6 (-3) open tickets

JIRA

# of tickets Δ State
128 -3 Open
21 -1 In Progress
9 +2 Ready for Testing
2 +2 Ready for Release

OSG Software Team

  • Update your Y7 effort numbers
  • OSG 3.4.17 packages not yet RFT: htcondor-ce, osg-test, xrootd-multiuser
  • The following hosts are uploading "Batch" (i.e. pilot) Gratia records and are not registered in the OSG topology
    • osg-gw-6.t2.ucsd.edu-COMET
    • aragon.cyverse.org
    • fsurf.osgconnect.net (should be uploading Payload records?)
    • sugwg-scitokens.phy.syr.edu
    • login03.osgconnect.net (should be uploading Payload records?)
  • Doc focus remaining tickets:
    • Document usage and examples of topology scripts (TimT)
    • Document topology deployment configuration and instructions (Derek)
    • Enable XRootD over HTTP (Edgar)
    • Review the topology registration doc (BrianL)
    • Review HTCondor-CE submission doc (Suchandra)
  • fetch-crl EL7 failures in the nightlies

Discussion

  • BrianL (AI)
    • Tag HTCondor-CE release, Mat to build
    • Talk to Mats about Cyverse
    • Pick a victim for fetch-crl EL7 test failures
  • Derek (AI)
    • Build xrootd-multiuser today, contact Bockjoo for testing
    • Working with Duncan to register the Syracuse CE
  • Suchandra (AI)
    • PR for osg-test packaging changes today
    • PR for HTCondor-CE submission doc review tomorrow
  • TimT (AI): Update topology script documentation along with the release

Support Update

  • Arizona State (BrianL, Mat, Suchandra): ironing out the appropriate setup for job submission and organizing a phone call
  • Harvard (BrianL): MCORE jobs aren't getting routed
  • Topology support (BrianL, Carl, Mat): miscellaneous contact and topology updates
  • UConn (Suchandra): ongoing issue with getting pilots
  • New Hosted CEs (Suchandra): New Mexico State University and the University of South Florida with a main cluster is 8k cores. North Dakota State recently came back to test opportunistic usage load on their network. They plan to run tests for a month or two. Long-term they will be setting up a new cluster in December/January to run opportunistically.

OSG Release Team

3.4.17 Δ Status
5 -4 Open
2 +2 In Progress
7 +5 Ready for Testing
0 +0 Ready for Release
12 +3 Total
  • OSG 3.4.17
    • Testing
      • Singularity 2.6.0
      • Gratia SLURM probe
      • cvmfs-x509-helper leaks 2 file descriptors per fetch
      • HTCondor 8.6.12
      • HTCondor 8.7.9
      • Pegasus 4.8.3
      • xrootd-lcmaps 1.40
    • Ready for Release
      • Nothing yet
  • Data
    • Testing
      • Nothing yet
    • Ready for Release
      • Nothing yet
  • GOC
    • Testing
      • Drop mirror.batlab.org from mirror list
    • Ready for Release
      • gracc-summary
      • gracc-archive
      • gracc-request

Discussion

  • BrianL (AI): Help find cvmfs-x509-helper testers
  • Edgar (AI): Madison ITB site still not getting GlideIns

OSG Investigations Team

In Progress:

  • XRootD documentation and docker overhaul is ongoing. Organized in StashCache in collaboration with Slate and NRP (PRP) collegues.
  • Discussions with with Internet2 to place StashCache in connection points.
  • Shepherding the xrootd plugins through the release process.
  • (Still ongoing) The repo server is going to move, once again. But there should be no noticable differences. If there is, then we failed.

Ongoing

Discussions

Derek (AI): Verify rsync service functionality for JINR mirror