Skip to Content.
Sympa Menu

interfed - Re: [inc-interfed] June 4 agenda / May 28 notes

Subject: Interfederation

List archive

Re: [inc-interfed] June 4 agenda / May 28 notes


Chronological Thread 
  • From: Ian Young <>
  • To:
  • Subject: Re: [inc-interfed] June 4 agenda / May 28 notes
  • Date: Tue, 11 Jun 2013 16:52:25 +0100
  • Authentication-results: sfpop-ironport01.merit.edu; dkim=pass (signature verified [TEST])


On 8 Jun 2013, at 17:31, Tom Scavo
<>
wrote:

> If InCommon is going to make another aggregate
> available to the community, we need to be able to bring some extra
> value to the table, beyond what a mindless aggregation operation
> provides. I'm still coming to grips with that.

You can be fairly mindless and still provide value. Obviously you don't need
to stop at that point in the longer term.

> Put another way, I would much prefer that eduGAIN pulled our metadata
> into their aggregate than the other way 'round.

Genuine participation in eduGAIN (or in any other kind of metadata exchange)
requires both. That's why we call them exchanges.

There is, however, no requirement for eduGAIN members to necessarily do both,
I made sure that was clarified in the new Declaration. So presenting eduGAIN
with metadata but not republishing anything from eduGAIN is permitted.
Obviously it's less beneficial than using both directions.

> That should be our short-term goal.

I don't see the need to pick one direction over the other; doing that will
just extend the time before you get the full benefits of the arrangement. It
should be possible to work on both at the same time.

>> P.S. plus, incidentally, eduGAIN isn't resourced to have individual end
>> entities all pulling aggregates from it. It's resourced for a relatively
>> small number of federation-sized participants.
>
> That's discouraging.

It may be, if you were hoping to point potentially every entity in InCommon
at http://mds.edugain.org. That's definitely not the intended use of the
service, however, and it has been sized accordingly.

> Does the server support HTTP compression?

I just checked to make sure, and it does not. It probably didn't occur to
them that it might be useful given the number of clients they were talking
about. Remember that this isn't a static file they are serving, it's
implemented as dynamic content. So if you don't care about the bandwidth or
the latency, compression is just an extra CPU cost.

-- Ian



Attachment: smime.p7s
Description: S/MIME cryptographic signature




Archive powered by MHonArc 2.6.16.

Top of Page