Skip to content

OSG Technology Area Meeting, 17 June 2019

Coordinates: Conference: 719-284-5267, PIN: 57363;
Attending: BrianL, Carl, Diego, Edgar, Marco Mambelli, Marian, Mat, TimT


BrianL at UChicago Wednesday, NYU next week, and vacation the following week. Returning July 5.

Triage Duty

  • This week: BrianL
  • Next week: Carl
  • 4 (+2) open FreshDesk tickets
  • 1 (+0) open GGUS ticket


# of tickets Δ State
155 -1 Open
29 +3 In Progress
5 -3 Ready for Testing
0 -2 Ready for Release

OSG Software Team

  • XCache:
    • AI (BrianL): Get ATLAS/CMS XCache RPM packaging in this week
    • AI (BrianL): Add XCache integration tests to the container CI
  • Hosted-CE:
    • Initial script for managing WN software/data from the CE host in the Operation team's hands
    • AI (Mat): execute remote host management plan
  • Blahp unification: Still working through HTCondor build issues.
  • AI (Derek): Review Gratia probe PRs (
  • AI (Mat): Waiting on Benedikt to verify access to the GLOW origin server
  • AI( BrianL, TimT): Madison ITB currently broken due to a pool password misconfiguration
  • Next doc focus July 25 (1:30-5:30pm CDT)


  • Edgar is also working on setting up a GlideinWMS frontend for BNL. This may be a Kubernetes setup or just a regular old VM.
  • Due to commas in some certificate DNs (looking at you, UCSD) this breaks the ability to run condor_off from GlideinWMS frontends against pilot startds. Marco has potential fixes involving using condor_mapfile instead of GSI_DAEMON_NAME.

Support Update

  • GLOW (BrianL): Some user jobs are being held in IIT startd's with the message "missing http plugin"
  • Let's Encrypt (Mat): A site asked for Let's Encrypt instructions for EL6. We don't have Let's Encrypt instructions for EL6 we know how to do this, it just needs to be documented
  • SWT2 (Carl): We've got a Gratia probe pull request in for a Slurm v18 fix; we're also investigating a strange dependency on the OSG user-vo-map that shouldn't be there anymore
  • Syracuse (BrianL): troubleshooting why their condor pool negotiator isn't matching their whole node jobs. Working with the Flightworthy team.

OSG Release Team

3.4.32 Δ Status
8 +8 Open
10 +10 In Progress
4 +4 Ready for Testing
0 +0 Ready for Release
22 +22 Total
  • OSG 3.4.32
    • Ready for Testing
      • cctools 7.0.14
      • XRootD 4.10.0-rc1
      • xrootd-scitokens 1.0.0
      • scitokens-cpp
    • Ready for Release
      • Nothing
  • Data
    • Nothing
  • Operations
    • Nothing
  • Contrib
    • Nothing
3.4.31 Δ Status
0 -7 Open
0 -6 In Progress
0 -7 Ready for Testing
7 +5 Ready for Release
7 -15 Total
  • OSG 3.4.31
    • Punted to 3.4.32
      • cctools 7.0.13
      • XRootD 4.10.0-rc1
      • xrootd-scitokens 1.0.0
      • scitokens-cpp
    • Released
      • Singularity 3.2.1
      • GlideinWMS proxy renewal (for Xenon)
      • HTCondor 8.6 patch python packaging to match EPEL
      • Upcoming: HTCondor 8.8.3
      • VO Package v93


  • AI (TimT): Take over osg-sites mailing list for community testing, populate it with Topology admin contacts, and write an introductory email [IN PROGRESS]

OSG Investigations Team

  • Accounting audit is complete with Fred. Report sent out to Operations team, will send further along as needed.
  • DONE: Adapting LIGO style cvmfs + singularity. Deployed and working.
  • cvmfs-authz-helper (cvmfs-x509-helper) - Derek ran out of time to tag / build.
  • XRootD Workshop went well. Not many questions…
  • SAND meeting this week. I’m sure there will be action items from this.
  • SLURM Gratia probe replacement has become urgent, will be done this week.
  • GRACC accounting taskforce - Tasking!!!


  • Working with operations investigating OAuth with co-manage. Waiting on co-manage (Wisconsin?!?!)




None this week