From nobody Tue Feb 9 14:04:12 2021 Return-Path: X-Original-To: oauth@ietfa.amsl.com Delivered-To: oauth@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BB18A3A0DAE; Tue, 9 Feb 2021 14:04:09 -0800 (PST) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -2.097 X-Spam-Level: X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vcGntX4Zod_2; Tue, 9 Feb 2021 14:04:06 -0800 (PST) Received: from mail-ej1-x633.google.com (mail-ej1-x633.google.com [IPv6:2a00:1450:4864:20::633]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 244463A0DAC; Tue, 9 Feb 2021 14:04:05 -0800 (PST) Received: by mail-ej1-x633.google.com with SMTP id i8so80465ejc.7; Tue, 09 Feb 2021 14:04:05 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=jmi9DwYJwMrbWQ5xyrPCkTWBnQqVxoYKX23x4JrPX/4=; b=dG7DmCMaUL9mg3Xx20TZZEvdosv6zl1tmafiu4s8NuMRzB6uIJwX9GkxVPa0MgfeVV jBrmEAdyODnoZ5jQd/eAPZqSyA4naP+D4YvGbyBm7FIp+dOHYWipiRmk0L1aODDY+CSP AGzgcwE3/Puuf+/zkQM5hvhlfQCVahLyy6hERlPY6iQbJOQi++XBZ9zwwz7ya6eNIAgf BKYmMralh4aYaXC9//kGhkzrinaF4adIOBnAmfWz3tLD9Ie8F8bQ4Cj1mbWyP5wYKuVC ohnJAw+F8L5czDkaCS71/00SDu531hz5kRzbl+0wYhl5+myKQyZfwXSHsZ0KkW5iP63K tUiA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=jmi9DwYJwMrbWQ5xyrPCkTWBnQqVxoYKX23x4JrPX/4=; b=IGtpjRqtvlAHRrTU24ngwLGbfXNpYihx9OXZuDRa249BFxqYoVcI/5wXVyE7C4ytpM k+D3jjShV58GyGnhb1ZBIqCqswK+e+eJtihNTuvIcnZnxCo5MTm6l1VBq1qKGMwOnod9 hoLb3WSOYlX9zZ8gKIjSaX1wr0BC3X0YTRTQ/5o5bYNutSoSCXnXBebWwVdmv8akhqf3 nuNq+Wp7P5dimno6IyaWo8u10m8GW6n+9iasMTvlwZ0k5Pu4vRrkNl18lo+Pi3/MUtPZ vpLbkcd082X6kxfI+RHuoeo9Rkm9Kuf7zh1bSOhFNLEERivyYfVV7KRsb4ELtLIKaTak 6GFA== X-Gm-Message-State: AOAM533IF+LKK51RsyNC/5gxJaEFMqKEe/Xfz9RMnctJD72XEIt2n+ya ufuUl/2w8YNEAtnTE5PzOXLjVrwxcNl+pxxH1Kc= X-Google-Smtp-Source: ABdhPJwlzXuoNhCNxgRM9faM7A4Qnwv/5ikD+SkMxm2VgNCl4U2t8eQ6Ti2McKn5Hr4GvTIN93NzOVqwAVQOdeHSVE4= X-Received: by 2002:a17:906:259a:: with SMTP id m26mr24986531ejb.399.1612908244237; Tue, 09 Feb 2021 14:04:04 -0800 (PST) MIME-Version: 1.0 References: In-Reply-To: From: Andrii Deinega Date: Tue, 9 Feb 2021 14:03:53 -0800 Message-ID: To: Neil Madden Cc: oauth , draft-ietf-oauth-jwt-introspection-response@ietf.org Content-Type: multipart/alternative; boundary="000000000000f4002c05baee7606" Archived-At: Subject: Re: [OAUTH-WG] JWT Response for OAuth Token Introspection and nonce X-BeenThere: oauth@ietf.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: OAUTH WG List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 09 Feb 2021 22:04:11 -0000 --000000000000f4002c05baee7606 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable I still don't see how your #1 and #3 points mitigate the replay attack when an attacker somehow eavesdrops a successful response from an AS (yes, it's signed by a public key) and then starts to replay it for other requests from the same client. The main problem here is that the client doesn't have a way to correlate the introspection response with the initial introspection request. Regarding #2, it's true that there are many proxies that do this and that. In practice, you don't always have control over the infrastructure where you may run your AS as I was saying before. Regards, Andrii On Tue, Feb 9, 2021 at 1:30 PM Neil Madden wrote: > Three points: > > 1. In many cases the JWT will be verified using a public key fetched over > the same TLS channel. > > 2. Many proxies can now also produce and consume JWTs for downstream > services, so end-to-end JWT is no more guaranteed than end-to-end TLS. > > 3. The JWT response already contains an iat claim which is sufficient to > judge freshness. > > It would be better to concentrate on ensuring end-to-end TLS rather than > trying to reinvent the same mechanisms in JWT form on top. > > =E2=80=94 Neil > > On 9 Feb 2021, at 20:38, Andrii Deinega wrote: > > =EF=BB=BF > How can you guarantee that there are always direct TLS connections betwee= n > a client and an AS hosted say some cloud provider where you have a little > control on their infrastructure? > > Even without all those cloud providers, how can you guarantee the same > when there are a bunch of different (software and hardware) components th= at > legitimately perform SSL offloading / DPI in front of an AS... or the > client may just use the proxy server? > > Regards, > Andrii > > On Tue, Feb 9, 2021 at 12:43 AM Neil Madden > wrote: > >> On 9 Feb 2021, at 06:55, Andrii Deinega wrote= : >> >> >> =EF=BB=BF >> Hi WG, >> >> I wonder if there are any particular reasons to not make nonce a >> mandatory parameter for the current JWT Response for OAuth Token >> Introspection draft. Or, at least, force an AS to include the nonce clai= m >> in a JWT response when nonce is presented in the introspection request >> similar to what happens with the similar scenario in the OpenID Connect = ID >> Token? >> >> >> https://openid.net/specs/openid-connect-core-1_0.html#:~:text=3DIf%20pre= sent%20in%20the%20Authentication%20Request%2C,value%20sent%20in%20the%20Aut= hentication%20Request. >> >> This will allow to mitigate replay attacks because clients can correlate >> the response with the initial request >> >> >> ID tokens involve flows using an insecure channel (the browser). This is >> not the case for introspection requests which happen over a direct TLS >> connection and so are already protected against replay attacks. >> >> =E2=80=94 Neil >> >> ForgeRock values your Privacy > > > ForgeRock values your Privacy --000000000000f4002c05baee7606 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
I still don't see how your #1 and #3 points mitigate t= he replay attack when an attacker somehow eavesdrops a successful response = from an AS (yes, it's signed by a public key) and then starts to replay= it for other requests from the same client.

The main problem here i= s that the client doesn't have a way to correlate the introspection res= ponse with the initial introspection request.

Regarding #2, it's= true that there are many proxies that do this and that. In practice, you d= on't always have control over the infrastructure where you may run your= AS as I was saying before.

Regards,
Andri= i


On Tue, Feb 9, 2021 at 1:30 PM Neil Madden <neil.madden@forgerock.com> wr= ote:
Three points:

1. In many cases the JWT will be verified using a public key fe= tched over the same TLS channel.=C2=A0

2. Many proxies can now also produce and consume JWTs for down= stream services, so end-to-end JWT is no more guaranteed than end-to-end TL= S.

3. The JWT response alr= eady contains an iat claim which is sufficient to judge freshness.=C2=A0

It would be better to concen= trate on ensuring end-to-end TLS rather than trying to reinvent the same me= chanisms in JWT form on top.=C2=A0

=E2=80=94 Neil

On 9 Feb 2021, at 20:38, Andrii Deinega <andrii.deinega@gmail.com> wrote:
=
=EF=BB=BF=
How can you guarantee that there are always direct TLS con= nections between a client and an AS hosted say some cloud provider where yo= u have a little control on their infrastructure?

Eve= n without all those cloud providers, how=C2=A0can you guarantee=C2=A0the sa= me when there are a bunch=C2=A0of different (software and hardware) compone= nts that legitimately perform SSL offloading / DPI in front of an AS...=C2= =A0 or the client may just use the proxy server?

<= div>Regards,
Andrii

On Tue, Feb 9, 2021 at 12:43 AM Ne= il Madden <neil.madden@forgerock.com> wrote:
On 9 Feb 2021= , at 06:55, Andrii Deinega <andrii.deinega@gmail.com> wrote:

=EF=BB=BF
Hi WG,

I wonder if t= here are any particular reasons to not make nonce a mandatory parameter for= the current JWT Response for OAuth Token Introspection draft. Or, at least= , force an AS to include the nonce claim in a JWT response when nonce is pr= esented in the introspection request similar to what happens with the simil= ar scenario in the OpenID Connect ID Token?

https://openid.net/specs/openid-connect-core-1_0.= html#:~:text=3DIf%20present%20in%20the%20Authentication%20Request%2C,value%= 20sent%20in%20the%20Authentication%20Request.

T= his will allow to mitigate replay attacks=C2=A0because clients can correlat= e the response with the initial request

ID tokens involve flows using an insecure channel (the browser). = This is not the case for introspection requests which happen over a direct = TLS connection and so are already protected against replay attacks.=C2=A0

=E2=80=94 Neil

ForgeRock values your Privacy

ForgeRock values your Privacy
--000000000000f4002c05baee7606--