Page tree
Skip to end of metadata
Go to start of metadata

Attributes for Collaboration and Federation

Call of Thursday, Nov. 9, 2017 

 Attending

  • Brad Christ, Southern Oregon University (chair)
  • Mark Scheible, MCNC 
  • Tom Barton, University of Chicago / Internet2
  • Brett Bieber, University of Nebraska
  • Ted Hanss, University of Michigan
  • Emily Eisbruch, InCommon/Internet2


 Working Group Resources

  1.  Wiki:  https://spaces.at.internet2.edu/x/ipiTBg
  2.  Charter:  https://spaces.at.internet2.edu/x/VgPABg
  3.  List:  attributes@incommon.org


Action Items from this call

[AI] (Brad) make a clone of the Steering R&S roadmap and modify it (Done) 

 [AI] (TomB) continue collecting data on users of R&S infrastructure in the US and report back to this working group.

[AI] (Brad) draft some common questions for outreach  around whether your institution participates in R&S and if not why not and reach out to NWAC on this (Done) 

[AI (Tom) will develop a selection of institutions for the outreach around R&S , for probing around  participation in R&S or lack of it and why  

DISCUSSION:

Review of 2017 TechEx Meeting feedback

Mark highlighted key points from the 2017 Tech Ex Attributes Working Group Meeting and consolidated some at the bottom of this Google Doc.  

Key points from the TechEx working group meeting:

  • There were comments at the Tech Ex meeting around identifying what has been preventing the release of R&S Attributes. That’s what this WG needs to understand and come up with mitigation.  
  • Communication is one way of increasing the number of institutions that release R&S.
  • Are research and scholarship different? One participant at the BOF said they don’t have researchers at their institution.  That institution participates in LIGO. Sometimes central IT does not know what researchers are doing on the campus.
  • Email is a key attribute for many research SPs.   
  • The best approach may bey to focus on communication, including how communication happens and who the message is communicated to. The person running the IDP may create attribute release policies for specific SPs and may not know how to implement R&S. May not know who to go to for approval, so they don’t implement R&S.  Perhaps we need to work with CIOs and perhaps we need to inform the CIOs know that NOT releasing R&S causes issues for the researchers.
  • Mark’s Example: “At MCNC, we have a community portal which contains reports and other information for our constituents. We recently upgraded to a new version of Drupal (with a new SP and entity ID). One of the universities could not get into the new portal because they didn’t have an attribute release policy (ARP) for the new SP.  They mentioned if the new SP had been R&S they could have gotten in without a problem.  It resulted in them needing to modify their ARP for the portal.”
  • At Virginia Tech they have found that a good policy is to release R&S across the board (much less work and maintenance) and to  implement consent (to deal with legal concerns).

Additional comments

  • Need to let people know that the risk is not perhaps as big as perceived. Can use consent to mitigate risk.  
  • It is disappointing the number of R1 institutions participating in R&S.   
  • If you release R&S, it is not to every SP your users access.  It is only being released to those SPs who have been approved for the R&S category.  But that is not always made clear.  Important to hone the message and get to the right audience.
  • Need to communicate to the right person, might not be the exec defined in InCommon.
  • When R&S was released it was not clear how much work was involved. The documentation was very technical, but perhaps hard to follow. Need simpler documentation. Eyes may glaze over the current documentation and institutions may say “I am not going to bother.”
  • Proposal/suggestion of enabling by default. (e.g. via the TIER Docker containers)
  • Use messaging to emphasize that R&S is a great benefit to faculty and staff. 
  • Provide tangible examples of the benefit
  • We may have been focusing too narrowly regarding R&S. 
  • Some audiences may not be familiar with federation and may need broader education.
  • The word "Metadata" can create red flags. Skip the word metadata perhaps in the messaging; Talk instead about directory data the way FERPA does, or specifically mention the R&S Attributes by (common) name,
  • We should talk about the library use case, and explain to librarians that R&S can address some of librarians' concerns, especially for inter-institution use of the library.

Summary and Next Steps

  • Roadmap (first item in charter)
  • Communication to right level
  • Education  (Good material in UC Davis thread on the InCommon Participants email list)
  • Simplify the doc (explain not a big deal technically)
  • Bundling
  • Use Shib to set default
  • 
After Roadmap is formed,  come up with policy recommendation to Steering
  • Steering has an attribute release roadmap doc  
  • It’s a table with tasks, it’s a plan, and we could revise it.
  • [AI] (Brad) make a clone of the Steering R&S roadmap and modify it (done)   

Should we update the charter for this working group?  https://spaces.at.internet2.edu/x/VgPABg
Decision: no need to revise the charter at this time

Pal Axelsson, author of GEANT assessment of impact of GDPR, published a paper yesterday.  TomB and Pal will be working on a blog on that topic.

TomB has been collecting data on users of Research and Science infrastructures in US. XSEDE 
(https://www.xsede.org/), Globus, etc.  Starting a good list.  

 [AI] (TomB) continue collecting data on users of R&S infrastructure in the US and report back to this working group.

 Outreach to CIOs and others around R&S

It was decided that members of this Attributes working group should speak to various size institutions who do not currently participate and find out why. Each of the WG members will make a few calls.

[AI] (Brad) draft some common questions for outreach  around whether your institution participates in R&S and if not why not and reach out to NWAC on this (Done) 

We should be sure we are reaching out to various size institutions. There is a grouping of the InCommon participants list by TIER. Tom can share that data.  

[AI (TomB) will develop a selection of institutions for the outreach around R&S, for probing around  participation in R&S or lack of it and why  


Next Call Thursday,. Dec 7, 2017
(Nov 23, 2017 cancelled due to Thanksgiving)

  • No labels