Re: [radext] 6614bis comment
Peter Deacon <peterd@iea-software.com> Fri, 21 April 2023 16:55 UTC
Return-Path: <peterd@iea-software.com>
X-Original-To: radext@ietfa.amsl.com
Delivered-To: radext@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DB978C14CE31 for <radext@ietfa.amsl.com>; Fri, 21 Apr 2023 09:55:35 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.498
X-Spam-Level:
X-Spam-Status: No, score=-1.498 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, KHOP_HELO_FCRDNS=0.399, RCVD_IN_ZEN_BLOCKED_OPENDNS=0.001, SPF_HELO_NONE=0.001, SPF_NONE=0.001] autolearn=no autolearn_force=no
Received: from mail.ietf.org ([50.223.129.194]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5R_4bGSFnRXB for <radext@ietfa.amsl.com>; Fri, 21 Apr 2023 09:55:35 -0700 (PDT)
Received: from aspen.iea-software.com (www.iea-software.com [70.89.142.193]) by ietfa.amsl.com (Postfix) with ESMTP id 4A231C14CF18 for <radext@ietf.org>; Fri, 21 Apr 2023 09:55:35 -0700 (PDT)
Received: from smurf (unverified [10.0.3.195]) by aspen.iea-software.com (Rockliffe SMTPRA 7.0.6) with ESMTP id <B0006186613@aspen.iea-software.com>; Fri, 21 Apr 2023 09:55:34 -0700
Date: Fri, 21 Apr 2023 09:55:39 -0700
From: Peter Deacon <peterd@iea-software.com>
To: Jan-Frederik Rieckers <rieckers@dfn.de>
cc: radext@ietf.org
In-Reply-To: <1652304a-4a3c-a098-1007-6d0045f6f508@dfn.de>
Message-ID: <e66b6fd8-aea6-4312-251b-9737928395db@iea-software.com>
References: <0c011fc2-16d0-324f-5d70-34df29d9d2bd@iea-software.com> <9eca79ad-7c4f-b34f-8964-7b4225a263a2@dfn.de> <132950AB-DB22-4496-A569-0CDBACB3F96B@deployingradius.com> <1652304a-4a3c-a098-1007-6d0045f6f508@dfn.de>
MIME-Version: 1.0
Content-Type: text/plain; CHARSET="US-ASCII"; format="flowed"
Content-ID: <36fc1b16-0e98-18f2-44d9-eced2eca4b7c@iea-software.com>
Archived-At: <https://mailarchive.ietf.org/arch/msg/radext/TgHCHt_-rZa3mGQ6if3r1Qa1PIw>
Subject: Re: [radext] 6614bis comment
X-BeenThere: radext@ietf.org
X-Mailman-Version: 2.1.39
Precedence: list
List-Id: RADIUS EXTensions working group discussion list <radext.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/radext>, <mailto:radext-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/radext/>
List-Post: <mailto:radext@ietf.org>
List-Help: <mailto:radext-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/radext>, <mailto:radext-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 21 Apr 2023 16:55:35 -0000
On Fri, 21 Apr 2023, Jan-Frederik Rieckers wrote: > On 20.04.23 15:04, Alan DeKok wrote: >>> I'm not that familiar with accounting since it is not really happening in >>> eduroam, so I have to read the RFCs to fully understand the problem, but >>> this is definitely a problem that should be addressed in the document. >>> Thanks for catching that. >> I would suggest replacing this text with a the Protocol-Error packet >> (RFC 7930). Something like: >> A server is usually configured to accept packets with a limited number of >> Code values. If a server receives a packet containing a Code outside of >> the acceptable list, it SHOULD reply with a Protocol-Error response >> containing an Error-Cause attribute with value "Unsupported Extension" >> [RFC5176]. A RADIUS/TLS client receiving such a response SHOULD log the >> error and stop sending packets of that Code value to the server >> That is generic enough to cover all Codes, too. > As far as I understood Peter's original comment the use case is the > following: > > A is connected to B. > B is connected to C and D. > C does not support accounting, D does. > A sends an Accounting-Update for C to B. > B proxies the request to C. > C does not support accounting and therefore rejects the update. > B marks the connection to C as "no accounting". > B proxies the reject message back to A. > A now erroneously marks the connection to B as "no accounting", although it > could reach D and D supports accounting. > If A has an accounting update for D, it will not send it, because the > connection to B is marked as "no accounting". > I guess the predicament is: Every request needs to have a response, but if we > proxy back the "I don't support accounting" response, clients may mark > proxies as "no accounting", although this is true for only one of their > connected servers. But if we just silently discard the "no accounting here" > response from the server, the accounting request remains unanswered. > (Peter, I hope I understood you correctly) Hi Janfred, Yes, there can be ambiguity with scope (e.g. peer vs realm). I like Alan's suggestion of the Protocol-Error code akin to SMPP's generic NAK indicating peer does not support the "extension". This way Error-Code 406 can continue to be received to indicate unsupported extension without risk of unintended global impact on the client. regards, Peter
- [radext] 6614bis comment Peter Deacon
- Re: [radext] 6614bis comment Jan-Frederik Rieckers
- Re: [radext] 6614bis comment Alan DeKok
- Re: [radext] 6614bis comment Jan-Frederik Rieckers
- Re: [radext] 6614bis comment Michael Richardson
- Re: [radext] 6614bis comment Peter Deacon
- Re: [radext] 6614bis comment Alan DeKok