OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

security-services message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [security-services] question on MNI request for SP Lite/IdP Lite


Replies inline.
::Ari
-----Original Message-----
From: Kyle Meadors [mailto:kyle@drummondgroup.com]
Sent: Wednesday, March 25, 2009 12:30 PM
To: security-services@lists.oasis-open.org
Subject: [security-services] question on MNI request for SP Lite/IdP Lite

For this year’s Liberty SAML IOP event, we are adding a new error test case for SP Lite and IdP Lite applications. Per the SAML conformance specification, SP Lite and IdP Lite MUST NOT support Name ID management messaging and functionality. As some of the vendor products allow their applications to be switched from SP to SP Lite mode (same with IdP and IdP Lite), we need to fully test this “switch” and verify they are properly acting in the SP Lite/IdP Lite mode.

 

To do this, we are expanding the error test tool. Once the ETT has established a SSO with the SP Lite/IdP Lite with a persistent name ID, we plan on generating a MNIRequest attempting to change the name ID. I have two questions which I would like feedback from SSTC on…
[Ari Kermaier] What business does a Lite implementation have using persistent NameIDs? My (admittedly dim) recollection of the origins of this conformance mode is that it stemmed from the concern of some SP implementers that they could not be conformant unless they had account linking data storage that would persist from one user session to the next. It seems to me, that the idea of a Lite implementation is inconsistent with the persistent NameID processing rules, and that differentiating based on support for MNI is really just a proxy for differentiating based on persistence.

 

1. When a SP Lite/IdP Lite receives a MNIRequest, what should be their proper response? Ignore the request? Generate an error <Status> response? Either is acceptable?
[Ari Kermaier] There should be no MNI service endpoint to contact for an implementation that doesn't support MNI. On the other hand, if the real requirement is that the provider does not support persistent NameIDs, then the MNI response should contain an error Status.

 

2. After generating the MNIRequest, I planned on verifying the SP Lite/IdP Lite did not accept the name ID change by sending a LogoutRequest using the new name ID from the MNIRequest message and looking for the error <Status> response that the SP Lite/IdP Lite does not recognize the principal. Then, send the LogoutRequest using the original name ID from the Response message and verifying principal is successfully logged out. Does this sound like a reasonable test scenario?

 

Thanks.

 

Kyle Meadors

Drummond Group Inc.

Principal, Test Process

(w) 615-212-0826

(c) 817-709-1627

kyle@drummondgroup.com

 



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]