From: Chad La Joie <lajoie@itumi.biz>
Date: November 23, 2010 12:39:54 PM EST
To: provision@internet2.edu
Subject: [provision] Ramping Up A Project, Or Not
Reply-To: provision@internet2.edu
Hey Everyone,
Tom Zeller and I spent some time at the fall I2 meeting discussing the topic of provisioning. We left with the feeling that we should consider spinning up a real project and he asked me if I'd take the technical lead, at least initially.
I believe we need to answer two initial questions before anyone decides there needs to be a real project.
1. Are current open source provisioning (e.g. Mule, Spring Integration, etc.) solutions adequate? If not, why not?
2. If existing solution are not adequate, what would be the scope of a provisioning project?
The above questions obviously raise the underlying question "what is adequate?". For the purpose of the initial discussions I would say an adequate solution is one where data from a source be provisioned to a target service, regardless of whether that target service is in the same administrative domain (i.e. it has to work in a federated world). I also believe that the requirement of working in a federated world brings with it the requirement to support one or more standard protocols that may be used to communicate data from source to target.
In order to help generate some discussion in this area I have asked a few people to join this list:
- Daniel Fisher, head of Middleware Services, at Virginia Tech
- Karsten Huneycutt who is currently working on a provisioning system for Univ of North Carolina
- Phil Smart, primary maintainer of the provisioning system at Cardiff University
And obviously I joined because I thought my experience in having developed a couple of provisioning solutions could be of use as well.
My plan, going forward, is to try to strike up some discussions, related to the above questions, on this mailing list. Once we have some data from those discussion I hope to arrange a couple of phone calls to discuss certain topics in more details. It should be obvious that the end goal is to be able to answer the discussion I posed above, and additional ones that are sure to come up.
So, to kick off the first discussion, I'd like to ask those people who have experience with particular solutions (commercial, open source, home grown, whatever) to provide some information about said solution and describe some of its best aspects and some of its shortcomings.
Thanks.
--
Chad La Joiehttp://itumi.biz
trusted identities, delivered