Interfederation

[Mark Scheible <>]

Hope to be able to join the call next week. Was traveling all day today. 

Mark

Sent from my iPhone

On Feb 5, 2013, at 11:27 AM, Jim Basney 
<>
 wrote:

> Hi,
> 
> Thanks everyone for another productive call today. My notes are below.
> Feel free to reply with corrections / additions.
> 
> For next Tuesday's call I propose the following agenda:
> 
> * interfed for K-12: Quilt/InCommon workshop outcomes (Mark)
> * interfed between InCommon & University systems
>  UTSystem+InCommon: current approach & future improvements (Paul)
>  UNL plans/requirements (Brett)
> * anything else?
> 
> [Assuming Mark, Paul & Brett can join us next Tuesday.]
> 
> Wiki link: https://spaces.internet2.edu/display/incinterfed
> 
> -Jim
> 
> -----
> 
> attending: JimB, TomS, MariaS, ScottC, ScottK, PaulC, IanY, SteveC
> new participants from last week:
>  Maria Schuett - Capella University
>    joined InCommon recently for
>      2 InCommon SPs: Simplicity & Digital Measures
>    interested to understand how universities federate with other SPs
>    here to observe & participate
>  Paul Caskey - UTSystem
>    interop between University system and InC federation
>  Steve Carmody - Brown
>    instructional & research use cases
>    courses include students at Universities on other continents
>    research use cases include LHC Physics
> proposed LIGO pilot project:
>  access to main LIGO wiki SP recently added to InCommon metadata
>    applied to be R&S
>  requires ePPN. currently manually added to ACL on LIGO wiki.
>  support access by Cardiff University collaborators
>  some Cardiff folks already have LIGO credentials but other
>    collaborators aren't officially on LIGO project but could
>    benefit from LIGO wiki access using home credentials
>  can we do it without LIGO SP joining federation? not just a one-off?
>  no immediate deadline for this - have time to do it right
>  then Glasgow next...
>  friction points:
>    metadata exchange
>    contracts
>    attribute release - R&S today is InCommon local.
>  just each load metadata from external federation?
>    without joining each other's federations?
>    LIGO already done this with Canadian federation.
>  how do we define interfederation?
>    IdP-SP direct exchange of metadata
>    each IdP/SP loads metadata from separate federations
>      downside: participants need to know what feds each is in
>    federations exchange metadata for participants
>  UK interfederation trial [1][2]:
>    Cardiff is involved already - good for LIGO pilot use case
>    separate test metadata aggregate can include external entities.
>      includes one of Cardiff's IdPs
>    ePPN release is a sensitive issue in Europe
>      hope for some intl version of R&S in long term...
>        isn't there a refeds work item on this? yes. see [3]
>    need a metadata source for LIGO's SP
>    get Cardiff IdP metadata from (options - "half-way houses"):
>      Cardiff directly
>      UK production metadata
>      UK export aggregate
>    Irish federation publishing a separate "export" metadata aggregate
>    could use InCommon production metadata aggregate to start
>  InCommon would need to do some work to handle metadata aggregates.
>  does it help for LIGO to make a formal request to InCommon to
>  provide a new metadata aggregate that includes UK export aggregate?
>    so LIGO doesn't need to download UK federation metadata directly.
>    given different LOA for different metadata sources,
>      putting them in same file raises policy concerns.
>      InCommon would need to re-sign.
>  interest: ScottK, IanY, TomS, JimB
>  what do we want to learn from this?
>    logistical issues that InCommon Ops would face as we scale to more
>      federations (not just UK)
>    what questions might campuses have of InCommon on policy side
>      before they decide to rely on InCommon's interfed aggregate?
> proposed metadata pilot?
>  publishing a metadata aggregate for InCommon participants
>  re-sign and republish another federation's metadata
>  dynamic metadata lookups?
>  want a single source of trust for the metadata
>  refer to UK references / plans
>  if Brown were consuming, would want ability to filter via tags
>  ref: MDRPI specification supports carrying registration info
> Brown SP pilot? AI: Steve will check.
>  Spring semester course with French students.
>  Will look for UK example.
> proposed edugain pilot project:
>  policy documents being revised now. JohnK and JimB are observers.
> follow-up on InCommon/Quilt workshop next week.
> progress on UTexas / InCommon interfederation?
>  UT already setup on a publish/subscribe basis.
>  Paul and Tom have discussed possible improvements to current setup.
> follow example from InCommon Social ID WG:
>  social ID working group instantiated a pilot service (UTexas hosted).
>  looking for a volunteer to create InCommon+UK metadata as a pilot
>    LIGO (ScottK) could do this in April/May
> 
> [1] http://www.ukfederation.org.uk/content/Documents/InterfederationTrialFAQ
> [2] http://iay.org.uk/blog/2012/08/uk-federation-metadata-aggregation
> [3]
> https://refeds.terena.org/index.php/REFEDS_Planning_Documents_2013#Entity_Categories