Choose timezone
Your profile timezone:
DOMA / TPC Meeting
→
Europe/Zurich
Description
Topic: WLCG DOMA TPC Meeting
Join Zoom Meeting
https://cern.zoom.us/j/99836057922?pwd=ZFhWN3NpYi9oZmwvM3pIRE9zdzFnZz09
Meeting ID: 998 3605 7922
Passcode: 733660
One tap mobile
+41315280988,,99836057922# Switzerland
+41432107042,,99836057922# Switzerland
Dial by your location
+41 31 528 09 88 Switzerland
+41 43 210 70 42 Switzerland
+41 43 210 71 08 Switzerland
+33 1 7037 2246 France
+33 1 7037 9729 France
+33 1 8699 5831 France
Meeting ID: 998 3605 7922
Find your local number: https://cern.zoom.us/u/aeB4ArMgmT
Mihai has changed the code in gfal. He asked about if the x509 scitoken issuer code an be packaged in an rpm in EPEL to make it easier to install it. Brian says it's not worth it and the code is so small it can be added to gfal directly. To avoid doubles we agree the code can be maintained in gfal directly because it is the main use case anyway.
A second problem he reports is how to inspect the tokens in order to make sure the storage returns what it is asked for. Brian says this is an extra burden that is not necessary and what Mihai is currently doing building a map with the tokens and what they are for is already good and benefits from being generic. Doing something more specific inspecting the tokens would mean starting to take into account different types of tokens (macaroons vs JWT).
the gfal code for people to test is here https://dmc-repo.web.cern.ch/dmc-repo/devel-next/el7/x86_64
The next step is to integrate it in FTS.
Experiments
Alessandra reports ATLAS has put INFN-T1 in production, it's the first non dcache and non DPM type of storage enabled by ATLAS. After some networking tests for the NOTED project done by Mario using HTTP-TPC between pic and CERN we also tried CERN-PROD in production but the efficiency was heavily affected by the race condition reported by Petr over a month ago https://github.com/xrootd/xrootd/issues/1314. Brian says he committed a fix but the problem is that the fix is included only in xrootd5 and not xrootd4. Elvin is not present so Alessandra will add Brian and Wei to the thread discussingthe issue to decide which is the quickest route to solve this problem. Otherwise CERN needs to be put on hold as far as HTTP-TPC is concerned.
http protocol on xrootd
Brian reports that the code written to map groups as discussed at the token hackaton is now merged and will be released in xrootd5.1.
Wei recommends sites that haven't upgraded to xrootd5 yet they should wait for 5.1 because 5.0.3 is ok but 5.1 is better.
There are minutes attached to this event.
Show them.
