[glue-wg] Cloud extensions

[Sill, Alan]

> On Sep 1, 2015, at 9:39 AM, Florido Paganelli <florido.paganelli at hep.lu.se> wrote:
> 
>  think the question is whether GLUE managed to foster such
> interoperability or not. BDII is LDAP based and pros and cons have been
> discussed in this mailing list before. I believe OSG just wants to drop
> that technology, but doesn't really need Glue in general. The claim I've
> heard in meetings is that they explicitly do NOT want to migrate to
> GLUE2, so I guess they don't even need the schema.
> 
> As for interoperability, the only existing interoperable systems based
> on GLUE2 I know are dCache, BDII, ARC and GOCDB to some extent. However,
> GLUE2 does not define aggregation, which makes interoperability loosely
> defined.
> 
> I am following the discussions in a working group at CERN about the next
> generation information system (read: for CERN experiments) and GLUE2 is
> rarely mentioned there. Some of the concepts also fail to be defined in
> GLUE2 without a different interpretation of some of the attributes,
> which actually takes us far from integration.
> Remember the discussion on how to identify a data transfer protocol on
> this mailing list? We even failed to have an agreement here....

Again, OGF’s interest is in documenting, providing forums and vehicles for discussion of, and helping to the extent that community input and an organized method for writing up the results of a discussion can help in arriving at the best practices and methods in a given area, including standard specifications and schema when these are necessary.

If GLUE2 can help in this area, then we are very happy to continue to support the group in its efforts to document, evolve and support the approach.  Any given project can of course make its own choices and can of course even document and publish these as OGF informational, community practice or experimental documents. Such documents are extremely helpful to allow the rest of the community to understand a particular approach, and they can be submitted, put out for public comment and published without regard to the need to establish consensus.

We are very happy, though, when the results of such documents and efforts lead to an interoperable approach that can then be published as a further guideline to encourage common efforts in a given area.

If there are discussions about a new information system, the process that we worked so hard as a community to create for an open, public way to document and get comments on it is available, and I hope that people will make use of it. 

Let me (or Jens or Wolfgang or Andre or any of the folks in the GFSG) know if a new working group needs to be chartered, or if a community group could help to encourage cross-project discussion. That’s a good way, I think, to proceed with the emergence of new technologies.

Alan