Attending: Dan, Jens (chair+mins), Ste, Wenlong, Teng, Vip, Matt, Brian, Sam, Winnie, PeteG, Patrick, Luke

Apols: RobC who is out babysitting LHCb.

0. Operational and Upgradeorational and Documentational Blog Posts

   No operational issues reported.

   Upgrade issues: Vip asked about the process for upgrading disk pools from SL6 to CentOS7.  In the case of Oxford
   storage is besteffortish, so maybe just declare downtime.  It depends on the customer; ATLAS's jobs are generally
   more fragile wrt temporarily missing files than CMS's.  Even when upgrading a single disk server at a time, there
   is a nonzero probability that a job will want to pull in lots of files and picks one at random that is on your
   disk server.  The alternative is to drain it, but that is a huge overhead.

1. Events - ideas for submissions (if applicable), attendance, etc.

1.1 IRIS F2F in Edinburgh: 
https://indico.ph.qmul.ac.uk/indico/conferenceDisplay.py?confId=446

The link above is very populated because it's last year's; the new link has not appeared as of this writing.
This year's event is in Edinburgh, 24/25 March (see chat).  At least on the data side we should be able to
say we have made progress; several things have been expected like TCP FTS (on not just GridFTP) and multi-VO
Rucio.

In yesterday's TWG, data transfers were mentioned as a topic (by PeteC), and also Rucio.  Maybe guidance
for the communities - is the barrier lower for newer communities?  What happened in the SKA evaluation?
Brian says they used the RAL FTS (thus delaying the TPC support upgrade) but apparently switched.

1.2 Annual cloud workshop in London (www.cloud.ac.uk)

Abstract submission closes this Friday.  Lots of vendors will be there.  StackHPC aiming to submit lots of
stuff and Andrew asked (yesterday) about other submissions on behalf of IRIS.

2. Storage (and CPU) Accounting - proposal from EOSC to use APEL: sneak 
preview, almost available for comments: 
https://confluence.egi.eu/display/EOSCDOC/Accounting

See also https://twiki.cern.ch/twiki/bin/view/LCG/AccountingTaskForce

We have three consumers/use cases of/for accounting (broadly speaking):
1 WLCG, using the WLCG accounting format/process (and the monit, see below)
2 IRIS, as AAAI was on the original roadmap, the third A being accounting.
3 Everyone else, like EGI, EOSC, etc., international interoperation, providing data to APEL.

We clearly have to do 1, and have done 1; there is a TWG for 2, although it's current state is not
known to be in wavefunction collapse.  For 3, the EOSC roadmap is aligned with EGI, and is based
on the assumption of the use of StAR.

How well is StAR supported in DPM and StoRM - because it would presumably work out of the box in
dCache, with its use in EGI.  It used to work for DPM, but has not been tested recently.


3. AOB

PeteG mentioned the CERN monit-grafana thingy (see link in chat), and asked why there is no CMS data.  It
is not a UK problem, as there is no CMS data for any country.

Your minute taker forgot to include the link to last week's minutes.


From Me to Everyone:  10:12 AM
https://cloud.ac.uk/
From Daniel Traynor to Everyone:  10:19 AM
24/25 march
From gronbech to Everyone:  10:37 AM
https://monit-grafana.cern.ch/d/mHqFLAbik/wlcg-storage-space-accounting?orgId=20
From Brian Davies @UKRI-STFC-RAL to Everyone:  10:43 AM
https://twiki.cern.ch/twiki/bin/view/LCG/CRICDataValidationInstruction
From Luke to Everyone:  10:45 AM
https://wlcg-cric.cern.ch/core/rcsite/detail/UKI-SOUTHGRID-BRIS-HEP/ --> Permission denied
From Me to Everyone:  10:12 AM
https://cloud.ac.uk/
From Daniel Traynor to Everyone:  10:19 AM
24/25 march
From gronbech to Everyone:  10:37 AM
https://monit-grafana.cern.ch/d/mHqFLAbik/wlcg-storage-space-accounting?orgId=20
From Brian Davies @UKRI-STFC-RAL to Everyone:  10:43 AM
https://twiki.cern.ch/twiki/bin/view/LCG/CRICDataValidationInstruction
From Luke to Everyone:  10:45 AM
https://wlcg-cric.cern.ch/core/rcsite/detail/UKI-SOUTHGRID-BRIS-HEP/ --> Permission denied
From gronbech to Everyone:  10:46 AM
Hi Luke I can't see that page even when I loginto cern.
We're sorry but access to this page is restricted
You are not allowed to access this page: following permissions should be granted: ['core.view_site']
From Luke to Everyone:  10:47 AM
yes, nor can I ;)
I am requesting permissions right now