[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Consensus Call on RADEXT WG re-charter



Hi,

> This is what you said: "...a RadSec to RADIUS gateway involves
> terminating TCP, TLS, and forwarding the contents to RADIUS.  If this
> isn't RADIUS, then by the same standards, any HTTPS to HTTP proxy isn't
> doing HTTP."  A Diameter to RADIUS gateway involves terminating TCP (or
> SCTP), TLS, and forwarding the contents to RADIUS.

With the devil being in the detail: With RadSec to RADIUS, the incoming packet 
is already a RADIUS packet as soon as it is decoded from the TLS tunnel. 
There is no translation of the content.
With Diameter to RADIUS, non-trivial translations need to be done, some of 
which can't even be done at all right now (Accepts >4096 bytes payload, AVPs 
with >253 bytes). We have yet to see a Diameter translation agent that can do 
bijective translations from one to another and back again (which I personally 
see as the holy grail of compatibility). I doubt we will ever. With RadSec to 
RADIUS, this is a triviality.

Greetings,

Stefan Winter

> If you really think 
> that this is different from your HTTPS->HTTP example, perhaps it's
> because you are conveniently ignoring the fact that HTTP is only defined
> over TCP.  However, one could just as easily argue that HTTP != HTTPS,
> given the need for different port numbers...
>
> >> I don't think that I said anything about good or evil :-), but you
> >> make my point for me: IPv6 is hardly "backward compatible" w/IPv4 &
> >> that's not a problem.  The problem would arise if someone claimed
> >> that it was because, 'Well, you can build a gateway'.
> >
> >  The protocols being transported on IPv4 can also be
> > transported on IPv6.
>
> I really don't know how that is relevant.  Dropping down a layer or 2,
> the same protocols can be carried by both Ethernet & ATM.  Are you going
> to claim that they are somehow compatible (backward, forward or
> upside-down)?
>
> > Similarly, UDP transport + RADIUS application
> > protocol is... RADIUS.  TCP transport + RADIUS application protocol
> > is... RADIUS.
>
> I know that you have read RFC 2865 (probably more often than I), so you
> know that there is no such thing as a standard "RADIUS application
> protocol", defined w/o regard for the underlying transport.
>
> >  The only way to call RadSec something other than RADIUS is
> > if you say that FTP over IPv6 is not FTP.  After all, FTP was
> > for IPv4,
>
> I can't find any mention of IPv4 in RFC 959; it does mention TCP,
> however.  Would FTPoUDP also be FTP?  How would FTP & FTPoUDP
> interoperate?  These questions seem much more to the point, I think.
>
> > and IPv6 is a new-fangled transport... so FTP over
> > IPv6 is not FTP... it's "magic".
>
> One more time: the version of IP being used is transparent & therefore
> irrelevant to FTP.  UDP transport is forever enshrined in RFC 2865,
> however; maybe that's a bad thing, but it's true nonetheless.
>
> >  Alan DeKok.
> >
> > --
> > to unsubscribe send a message to
> > radiusext-request@ops.ietf.org with the word 'unsubscribe' in
> > a single line as the message text body.
> > archive: <http://psg.com/lists/radiusext/>
>
> --
> to unsubscribe send a message to radiusext-request@ops.ietf.org with
> the word 'unsubscribe' in a single line as the message text body.
> archive: <http://psg.com/lists/radiusext/>



-- 
Stefan WINTER

Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de 
la Recherche
Ingenieur Forschung & Entwicklung

6, rue Richard Coudenhove-Kalergi
L-1359 Luxembourg
E-Mail: stefan.winter@restena.lu     Tel.:     +352 424409-1
http://www.restena.lu                Fax:      +352 422473

Attachment: signature.asc
Description: This is a digitally signed message part.