Duncan, Winnie, Sam, Jens (chair+mins), Raja, Teng, JohnH

0. OBP

   Sam's writing a post...

   Operational issues, Raja mentions the dynafed-large-files-to-Echo-S3 problem which affects DUNE
   - it "should have been fixed" by the end of last year.

   dCache also needs to support non-GridFTP TPC to RAL, particularly since DUNE data comes from
   FNAL (and not from the target site).  However, as a fallback option RAL could also take DUNE
   data from CERN, or as a fallback fallback, from anywhere else.

1. DIRAC *can* be made to use Rucio in the sense that it can register files in Rucio, just
   like it can in DFC.

   Duncan points out there is a Rucio workshop in Oslo soonish.

   The background is how we "sell" GridPP-as-a-data-management-infrastructure to users of IRIS
   (it assumes that we want to do so, but the assumption is not entirely far fetched as several
   IRIS non-LHC experiments have already trialled Rucio, e.g. SKA and DUNE.  See also chatlog.)

2. Docs

   A quick reminder to people to check their Key Docs...!


[link to updated google doc for (CERN) Data Lakes]
Daniel
t2k (and i assume HyperK) using DFC.
I have talked to hyperK about using rucio in the long term.
time scale for t2K means DFC is the right choice
DP
Duncan
Indeed
DR
Today at 10:32 AM