shibboleth-dev - RE: AA implementation experiment
Subject: Shibboleth Developers
List archive
- From: Scott Cantor <>
- To: 'Shibboleth Design Team' <>
- Subject: RE: AA implementation experiment
- Date: Thu, 25 Apr 2002 17:28:56 -0400
- Importance: Normal
- Organization: The Ohio State University
> Well, how do you feel about these numbers?
I guess they're about what I expected. I'm not surprised reads were more
expensive, since that's mostly spent constructing objects. I'm sure it's
not the I/O.
I think it's sufficient for a small scale deployment, and the primary
concern is in getting code running that demonstrates the full feature
set, so if this is the easiest way to do that, I can live with it.
My biggest concern would probably be more about reliability than
performance. How does synchronization of the files between the AA's
responder and UI work? Does it lock the files somehow and "just work"? I
guess you'd back it up by archiving up all the files in the ARP storage
folder. Again, not something I'd want to deploy on a wide-scale or with
lots of user ARPs, just in case something went wrong.
-- Scott
------------------------------------------------------mace-shib-design-+
For list utilities, archives, subscribe, unsubscribe, etc. please visit the
ListProc web interface at
http://archives.internet2.edu/
------------------------------------------------------mace-shib-design--
- AA implementation experiment, Parviz Dousti, 04/25/2002
- RE: AA implementation experiment, Scott Cantor, 04/25/2002
Archive powered by MHonArc 2.6.16.