Weekly operations meeting for sites doing DUNE computing
No meeting on Dec 26 and Jan 2
Discussion on Kirby’s slide
Jonathan—don’t have people showing things that anyone could look up.
Peter—good idea with the hand raised and the formalism
Peter—Edinburgh is not not going to produce a slide every a week
Should concentrate on sites that have problems
Kirby—recognize concern
Want to make sure that the site is happy
How can we get positive confirmation that things are OK
Peter—why would you want a slide for sites that are just running jobs.
Certainly won’t find time
Grid-PP maybe
Elisabetta—share point of Peter that producing slide of every week is probably a bit too much.
Provide a “T0” slide for every site.. if there is a template of the site that is needed.
Then every week if site has an issue it can prepare a slide.
Jonathan—for those sites or consortia of sites—“everything’s fine” can satisfy problem of
Driving engagement.
If people think participation is of value then they will participate.
Changes suggested for meeting otherwise are good
Steve—worst sites don’t show up or sometimes don’t have a contact at all
Should automate what we can.
——
EGI accounting looks
GGUS ticket @ CERN re. Rucio upload failing
Tomorrow Dec. 6 EOS downtime @ CERN—eospublic looking at VOMS attributes now.
NP02 status
Will turn off machines in EHN-1 next week
NP04 status
Coordination meeting was cancelled, last one on 13th.
Scale test of DAQ not likely.
Some trigger primitives written to disk…
Production—
Tejin, about to start production of single pion sample.. tried to add into production calendar.
Tejin has requested access
Ken—only running at NERSC and not is going to make a dent.
Elisabetta started production of CRP3 reconstruction.. waiting for feedback. Will start the full prod in Jan.
Production of 6m track is starting this week.
Data management ops
Sent some data to PIC.
Data dispatcher part of data challenge mostly complete.
————————
Site report—
Edinburgh Rucio cache plans at Edinburgh will discuss tomorrow
QMUL—network situation is improving.. now have core network and new firewalls in place.
100 Gbit link to be up and running before Christmas break
CCIN2P3—tomorrow scheduled shutdown 12 hrs reduced.
50TB available on dCache
10TB available on tape
Fermilab—
No dCache downtime before Christmas
Canada —jobs normal..
CCB meeting on Thursday 12/8