RE: : digest issue 11 closure (fwd)
Bernard Aboba <[email protected]>
| Newsgroups | gmane.ietf.aaa |
|---|---|
| Message-ID | <[email protected]> |
> Well, the RADEXT WG *does* have a requirement that RADIUS extensions be > interoperable with Diameter. I guess what you're saying is that the AAA > WG does *not* have a requirement that Diameter be interoperable with > RADIUS extensions. The original idea was for Diameter to offer a superset of RADIUS functionality, so that it would be possible to install Diameter agents and servers in a network with RADIUS-capable NAS devices. This was thought to be important because common NAS devices do not today support Diameter, and may not do so for a long time. At the same time, Diameter needed to be able to offer functionality such as improved transport and new applications that could not be built on top of RADIUS. Since RADIUS could not provide some of those features, interoperability was not required for those elements (such as failover, and Credit Control). > There are two SIP Digest Authentication protocols, one in Diameter and > one in RADIUS. My understanding is that one originated in 3GPP and the > other originated in 3GPP2. Are you suggesting that these are disjoint > solution sets that need not interoperate? As far as I can see SIP Digest authentication does not require use of features uniquely available in Diameter. Since this is a new application, one might posit that deployments will largely focus on Diameter rather than RADIUS, but that is a hypothesis which could be wrong. If in fact RADIUS Digest becomes popular, then it may become necessary to gateway RADIUS Digest to Diameter SIP, and without interoperability that could be hard. Why tempt fate?