LHC Computer Grid - Rollout
> [mailto:[log in to unmask]] On Behalf Of Gonzalo Merino
said:
> So, what you say is that for a classic SE the mount point is
> not taken
> from GlueSARoot, but from the GlueCESEBindCEAccesspoint
> published by the CE?
Yes, that was always true - the full storage path is the concatenation
of the access point and the saroot.
> By the way, what defines csatorgrid.pic.es as a Classic SE, from the
> information system point of view? Is it that the
> GlueSEName=<sitename>:disk?
Yes. Like many things it's a short-term hack which lasted a long time!
The new schema will have a new ControlProtocol object, similar to the
AccessProtocol, to replace that.
Stephen
|