[ ... ]
> 1) Amount of non-trivial storage used by non-WLCG, non-T2K VOs.
Well, we have 13T to 'pheno' (home team and 11TB of that is one
user), 1.5TB for supernemo and 1TB for 'biomed' (and many thanks
again to the author of 'dpm-sql-usage-by-vo-user' :->).
> 2) Of those VOs, any contact you've had with them (specifically
> regarding storage, but any grid contact you feel is relevant).
In general these have been about VOMS certificate changes,
involving mostly 'biomed' and 'enmr.eu'.
> 3) If you feel a particular VO would benefit from Storage Group
> services (including consistency checking).
'pheno', as they do occasionally use quite a bit of storage.
And I guess I have already mentioned one of their issues: they
send out 2,000-5,000 job batches, and it is very slow for them
to collect that many job output files from scattered sites via
LFC/SRM access (in part because the 'pheno' LFC has been slow,
but not just).
BTW I had a brief look into Chirp (suggested by someone) as
possible way for them to write all the outputs conveniently to
home base.
|