DOMA / TPC Meeting

Wednesday 8 May 2019, 17:30 → 18:30 Europe/Zurich

DOMA / TPC Meeting Notes

• Attending: Al, Andrea M, Brian B, Dmitriy L, Duncan R, Horst S, Katy E, Matt D, Paul M, Petr V, Andrea C., Nikola H.

• Brian’s updates:

  • With Xrootd 4.9.1 out, I’ve been bothering US sites (including MWT2) to enable dteam and get HTTPS configured.  Hopefully getting to the point where they will be able to join the smoke tests next week.

• Paul:

  • Smoke tests are being run regularly now and behaving -- ready to put them into a cronjob.

  • If we run things regularly, should we have some sort of history or monitoring?

    • Should we “game-ify” things to generate interest?

    • How does this overlap with Rucio + MONIT?  Is there an easy way to setup MONIT history?

    • Can we (should we?) integrate with downtimes?

    • Suggestion(Brian) : Let’s have Rucio be our long-term record and simply increment / decrement an integer every run.  Keep things simple.

  • Andrea C. notes that the smoke tests are broken on RHEL7.

    • ACTION(Brian): Brian has a patch, should send it as a PR.

  • CHEP abstracts:

    • ACTION(Brian): Will put together one about the DOMA TPC working group

    • ACTION(Paul): Will put together one about HTTP TPC.

  • Q from Brian: Can DESY start using HTTP for VO-specific (CMS, ATLAS) production/debug transfers?

    • Brian has a “one line” patch to PhEDEx for those wanting to test in CMS.

    • U Bonn has done this.

    • What about Prague?

    • Imperial maybe?

• Andrea M:

  • fts3.cern.ch should be ready to do HTTP & Xrootd TPC.

  • ATLAS FTS instance does not have the SciTokens issuer library.

  • How do we start enabling sites?  UBonn had an issue in that many sites advertised Xrootd support - but very few were functional in reality.

    • Can we ask ATLAS to “press the reset button” on Xrootd or HTTP TPC?  Basically disable it at all sites and enable things one-by-one we know things work.

    • Basically, sites are scared to enable HTTP/Xrootd at the destination because they don’t want to debug broken source sites.

    • Petr notes there isn’t monitoring for individual protocols yet in ATLAS.  This may pose operational problems.

    • ACTION(Petr): Bring this up internally in ATLAS.

    • We probably need a similar smoke test for Xrootd.

      • Looking for volunteer.

• RAL:

  • Made very good progress on Xrootd/ECHO.  Got at least the proof of concept working against Brunel + SLAC.

  • ACTION(Thomas): Enable RAL.

  • ACTION(Thomas): Andrea M reports that functional/stress tests stopped working.  Need to go revive things.

• OU:

  • Good progress also -- know all the package versions needed, will update and see what breaks!

  • Paul: Can you also enable the HTTP support?  Horst: Sure! Send instructions.

ACTION(Brian): Will send out a link to Horst.

17:30 → 17:50 Xrootd Protocol Update

Speaker: Wei Yang (SLAC National Accelerator Laboratory (US))

17:50 → 18:10 HTTP Protocol Update

Speaker: Brian Paul Bockelman (University of Nebraska Lincoln (US))

18:10 → 18:30 Discussion