Testbed Support for GridPP member institutes
> [mailto:[log in to unmask]] On Behalf Of Alessandra Forti said:
> first time we started to talk about this was when atlas
> screamed that if
> the storage is declared permanent they expect for the data to
> be recovered
> if something happens. I think it was in january/february.
I don't think that was about backups, it was the fact that some admins
were deleting files, or even scratching whole SEs and reinstalling them.
We still have that issue in the question of SRM migration. That's
something where in theory the "volatile" definition is appropriate, it
would let you scratch an SE without having to migrate the files - but
even if volatile did have some effect, it probably wouldn't be a good
argument to use it just in case you might want to migrate the SE at some
point.
> Perhaps it
> wasn't the right place however I thought it was the place
> where to start
> to put down points for a discussion with the experiments. If
> you have a
> better aproach I'm open to suggestions.
Even in the broader picture I don't think it's obvious where such
discussion happens, there are several bodies - NA4, ARDA, the GAG, the
BSWG, the GDB, the SCs - and I suspect that still a lot (most?) of the
interaction happens informally.
Stephen
|