Re: [dhcwg] I-D Action: draft-ietf-dhc-sedhcpv6-18.txt

"Bernie Volz (volz)" <volz@cisco.com> Sun, 18 December 2016 21:57 UTC

Return-Path: <volz@cisco.com>
X-Original-To: dhcwg@ietfa.amsl.com
Delivered-To: dhcwg@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DC144129450 for <dhcwg@ietfa.amsl.com>; Sun, 18 Dec 2016 13:57:28 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -17.621
X-Spam-Level:
X-Spam-Status: No, score=-17.621 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_HI=-5, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, RP_MATCHES_RCVD=-3.1, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001, USER_IN_DEF_DKIM_WL=-7.5] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=cisco.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 LbKYp_uGALyj for <dhcwg@ietfa.amsl.com>; Sun, 18 Dec 2016 13:57:22 -0800 (PST)
Received: from alln-iport-2.cisco.com (alln-iport-2.cisco.com [173.37.142.89]) (using TLSv1.2 with cipher DHE-RSA-SEED-SHA (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 57ED112944C for <dhcwg@ietf.org>; Sun, 18 Dec 2016 13:57:22 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=cisco.com; i=@cisco.com; l=405914; q=dns/txt; s=iport; t=1482098242; x=1483307842; h=from:to:subject:date:message-id:references:in-reply-to: mime-version; bh=SinsJFdFmcFfDy83kzA/rx/pif+3oEpoLCaRNLxIhCM=; b=NGY52AZQEJ8GmrFU8yGBObeUkSAvnHQYdgjj2Ywq9GAOH8i4Q0B538G1 pPadeUfdQeheGowIiVshH9f5TS4g6DC98/smfzlEAc5+/kLATmp+fxSXm jA+OOjt4j2KJ16H6PKkRj9kyYb8iuXbWWrRJHlIua/IEqLyP/BsyspDHb k=;
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: A0CmBACNBVdY/4MNJK2ERAEBAQEBxkwEAgECAYEK
X-IronPort-AV: E=Sophos;i="5.33,371,1477958400"; d="scan'208,217";a="361133971"
Received: from alln-core-1.cisco.com ([173.36.13.131]) by alln-iport-2.cisco.com with ESMTP/TLS/DHE-RSA-AES256-SHA; 18 Dec 2016 21:57:21 +0000
Received: from XCH-RCD-001.cisco.com (xch-rcd-001.cisco.com [173.37.102.11]) by alln-core-1.cisco.com (8.14.5/8.14.5) with ESMTP id uBILvLC6010318 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=FAIL); Sun, 18 Dec 2016 21:57:21 GMT
Received: from xch-aln-003.cisco.com (173.36.7.13) by XCH-RCD-001.cisco.com (173.37.102.11) with Microsoft SMTP Server (TLS) id 15.0.1210.3; Sun, 18 Dec 2016 15:57:20 -0600
Received: from xch-aln-003.cisco.com ([173.36.7.13]) by XCH-ALN-003.cisco.com ([173.36.7.13]) with mapi id 15.00.1210.000; Sun, 18 Dec 2016 15:57:20 -0600
From: "Bernie Volz (volz)" <volz@cisco.com>
To: Lishan Li <lilishan48@gmail.com>, dhcwg <dhcwg@ietf.org>
Thread-Topic: [dhcwg] I-D Action: draft-ietf-dhc-sedhcpv6-18.txt
Thread-Index: AQHSTs4GXMOx5uePjkC6PykSIXEqwqD56A0AgBRbwzA=
Date: Sun, 18 Dec 2016 21:57:19 +0000
Message-ID: <352fe8719ab34fd3ad7b0af2e7127bb8@XCH-ALN-003.cisco.com>
References: <148092498114.3294.14134801279747342720.idtracker@ietfa.amsl.com> <CAJ3w4NfufMn9yzACowX728Cxxn4Rdci5pXAwVYB3d2fKDtNF2Q@mail.gmail.com>
In-Reply-To: <CAJ3w4NfufMn9yzACowX728Cxxn4Rdci5pXAwVYB3d2fKDtNF2Q@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach:
X-MS-TNEF-Correlator:
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [10.98.1.196]
Content-Type: multipart/alternative; boundary="_000_352fe8719ab34fd3ad7b0af2e7127bb8XCHALN003ciscocom_"
MIME-Version: 1.0
Archived-At: <https://mailarchive.ietf.org/arch/msg/dhcwg/Kb8yi-jPKRrHZZV0kIojm2wAYNs>
X-Mailman-Approved-At: Sun, 18 Dec 2016 13:58:18 -0800
Subject: Re: [dhcwg] I-D Action: draft-ietf-dhc-sedhcpv6-18.txt
X-BeenThere: dhcwg@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: <dhcwg.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/dhcwg>, <mailto:dhcwg-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/dhcwg/>
List-Post: <mailto:dhcwg@ietf.org>
List-Help: <mailto:dhcwg-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/dhcwg>, <mailto:dhcwg-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 18 Dec 2016 21:57:29 -0000

Hi:

I read through this draft and do have extensive comments. While there are a lot of improvements, I think some work is needed on the document. Please note that I still have to re-review some of section 6 and 7, but as there were plenty of comments already, I figured it best to provide you these earlier than waiting to finish (as it may also be a bit difficult to finish that review over the next several weeks because of the holidays).

I’ve included much of the draft below as that makes it easier to comment on the issues. My comments should start with BV>.

Let me first raise a few points to consider:


-          A server could have multiple algorithms and (possibly multiple certificates) and so when a server receives a client’s Encrypted-Query message how does it decrypt it (choose from its algorithms or certificates)? If it tries several, how can it tell when it is successful? I guess by verifying the signature? But this also requires the server to decrypt the packet and look for this option – and the decrypted packet could be complete junk (which the server obviously should deal with but there is a risk that this junk could cause the server to crash)? I’m not sure if the Encryption Key Tag was supposed to provide for this as (commented below) there are no details on how this is generated (section 10.1.5 does not indicate anything useful). Also, having the server have to try to decrypt several times seems a bit costly to me. (Again, the server cannot even tell who the client is so there’s no easy way for the server even to remember this and associate it with something in the received packet.) I wonder whether there should be some mechanism for the server to provide some “encryption-id” to the client (in the Encrypted-Response) which the client then sends in an option in the Encrypted-Query so the server knows how to decrypt the message?


-          Is Trust on First Use (TOFU) supported? It didn’t seem so? (Section 6 says to “or decide[d] to drop the message”). That may be fine and perhaps it was dropped in recent discussions (I had not followed all of them carefully). Or perhaps that will be for some follow on work.


                            Secure DHCPv6
                       draft-ietf-dhc-sedhcpv6-18

Abstract

   DHCPv6 includes no deployable security mechanism that can protect
   end-to-end communication between DHCP clients and servers.  This
   document describes a mechanism for using public key cryptography to
   provide such security.  The mechanism provides encryption in all
   cases, and can be used for authentication based on pre-sharing of
   authorized certificates.


Jiang, et al.             Expires June 7, 2017                  [Page 2]

Internet-Draft                  SeDHCPv6                   December 2016


1.  Introduction

   The Dynamic Host Configuration Protocol for IPv6 (DHCPv6, [RFC3315])
   allows DHCPv6 servers to flexibly provide addressing and other
   configuration information relating to local network infrastructure to
   DHCP clients.  The protocol provides no deployable security
   mechanism, and consequently is vulnerable to various attacks.

   This document provides a brief summary of the security
   vulnerabilities of the DHCPv6 protocol and then describes a new
   extension to the protocol that provides two additional types of
   security:

   o  authentication of the DHCPv6 client and the DHCPv6 server to
      defend against active attacks, such as spoofing.

   o  encryption between the DHCPv6 client and the DHCPv6 server in
      order to protect the DHCPv6 communication from pervasive
      monitoring.

   The extension specified in this document applies only to end-to-end
   communication between DHCP servers and clients.  Options added by
   relay agents in Relay-Forward messages, and options other than the
   client message in Relay-Reply messages sent by DHCP servers, are not
   protected.  Such communications are already protected using the
   mechanism described in section 21.1 in [RFC3315].

   This extension introduces two new DHCPv6 messages: the Encrypted-
   Query and the Encrypted-Response messages.  It defines six new DHCPv6
   options: the Algorithm, Certificate, Signature, Increasing-number,
   Encryption Key Tag option and Encrypted-message options.  The
   Algorithm, Certificate, Signature, and Increasing-number options are
   used for authentication.  The Encryption-Query message, Encryption-
   Response message, Encrypted-message option and Encryption Key Tag
   option are used for encryption.

BV> To be consistent, should “Encryption Key Tag” be changed to “Encryption-Key-Tag” throughout?

2.  Requirements Language and Terminology

BV> Shouldn’t this just be Requirements Language as section 3 is Terminology? Or, perhaps combine these sections into one?

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC2119] when they
   appear in ALL CAPS.  When these words are not in ALL CAPS (such as
   "should" or "Should"), they have their usual English meanings, and
   are not to be interpreted as [RFC2119] key words.



Jiang, et al.             Expires June 7, 2017                  [Page 3]
Internet-Draft                  SeDHCPv6                   December 2016


3.  Terminology

   This section defines terminology specific to secure DHCPv6 used in
   this document.

   secure DHCPv6 client:  A node that initiates a DHCPv6 request on a
                   link to obtain DHCPv6 configuration parameters from
                   one or more DHCPv6 servers using the encryption and
                   optional authentication mechanisms defined in this
                   document.

   secure DHCPv6 server:  A DHCPv6 server that implements the
                   authentication and encryption mechanisms defined in
                   this document, and is configured to use them.

BV> There are some other abbreviations used in this document that are not “defined” on first use and perhaps best added here? For example, PKI and CA (there may be others). Or, do you want to pull in terminology from a security document (RFC)?

4.  Security Issues of DHCPv6

   [RFC3315] defines an authentication mechanism with integrity
   protection.  This mechanism uses a symmetric key that is shared by
   the client and server for authentication.  It does not provide any
   key distribution mechanism.
BV> Not sure if it matters, but this was deprecated in 3315bis?

   For this approach, operators can set up a key database for both
   servers and clients from which the client obtains a key before
   running DHCPv6.  However, manual key distribution runs counter to the
   goal of minimizing the configuration data needed at each host.
   Consequently, there are no known deployments of this security
   mechanism.

   [RFC3315] provides an additional mechanism for preventing off-network
   timing attacks using the Reconfigure message: the Reconfigure Key
   authentication method.  However, this method protects only the
   Reconfigure message.  The key is transmitted in plaintext to the
   client in earlier exchanges and so this method is vulnerable to on-
   path active attacks.

   Anonymity Profile for DHCP Clients [RFC7844] explains how to generate
   DHCPv4 or DHCPv6 requests that minimize the disclosure of identifying
   information.  However, the anonymity profile limits the use of the
   certain options.  It also cannot anticipate new options that may
   contain private information is defined.  In addition, the anonymity
BV> Drop “is defined”?
   profile does not work in cases where the client wants to maintain
   anonymity from eavesdroppers but must identify itself to the DHCP
   server with which it intends to communicate.

   Privacy consideration for DHCPv6 [RFC7824] presents an analysis of
   the privacy issues associated with the use of DHCPv6 by Internet
   users.  No solutions are presented.



Jiang, et al.             Expires June 7, 2017                  [Page 4]

Internet-Draft                  SeDHCPv6                   December 2016


   Current DHCPv6 messages are still transmitted in cleartext and the
   privacy information within the DHCPv6 message is not protected from
   passive attack, such as pervasive monitoring [RFC7258].  The privacy
   information of the IPv6 host, such as DUID, may be gleaned to find
   location information, previous visited networks and so on.  [RFC7258]
   claims that pervasive monitoring should be mitigated in the design of
   IETF protocol, where possible.

   To better address the problem of passive monitoring and to achieve
   authentication without requiring a symmetric key distribution
   solution for DHCP, this document defines an asymmetric key
   authentication and encryption mechanism.  This protects against both
   active attacks, such as spoofing, and passive attacks, such as
   pervasive monitoring.

5.  Secure DHCPv6 Overview

5.1.  Solution Overview

   The following figure illustrates secure DHCPv6 procedure.  Briefly,
BV> add the (illustrates the secure)
   this extension establishes the server's identity with an anonymous
   Information-Request exchange.  Once the server's identity has been
   established, the client may either choose to communicate with the
   server or not.  Not communicating with an unknown server avoids
   revealing private information, but if there is no known server on a
   particular link, the client will be unable to communicate with a DHCP
   server.

   If the client chooses to communicate with the selected server(s), it
   uses the Encrypted-Query message to encapsulate its communications to
   the DHCP server.  The server responds with Encrypted-Response
   messages.  Normal DHCP messages are encapsulated in these two new
   messages using the new defined Encrypted-message option.  Besides the
   Encrypted-message option, the Signature option is defined to verify
   the integrity of the DHCPv6 messages and then authentication of
BV> Then should be the?
   client and server.  The Increasing number option is defined to detect
   replay attack.
BV> a replay attack (or replay attacks)?


Jiang, et al.             Expires June 7, 2017                  [Page 5]
Internet-Draft                  SeDHCPv6                   December 2016


           +-------------+                           +-------------+
           |DHCPv6 Client|                           |DHCPv6 Server|
           +-------------+                           +-------------+
                  |            Information-request           |
                  |----------------------------------------->|
                  |             Algorithm option             |
                  |           Option Request option          |
                 |                                          |
                  |                    Reply                 |
                  |<-----------------------------------------|
                  |             Certificate option           |
                  |             Signature option             |
                  |          Increasing-number option        |
                  |         Server Identifier option         |
                  |                                          |
                  |            Encryption-Query              |
                  |----------------------------------------->|
                  |          Encrypted-message option        |
                  |          Server Identifier option        |
                  |         Encryption Key Tag option        |
                  |                                          |
                  |            Encryption-Response           |
                  |<-----------------------------------------|
                  |          Encrypted-message option        |
                  |                                          |

                     Figure 1: Secure DHCPv6 Procedure

5.2.  New Components

   The new components of the mechanism specified in this document are as
   follows:

   o  Servers and clients that use certificates first generate a public/
      private key pair and then obtain a certificate that signs the
      public key.  The Certificate option is defined to carry the
      certificate of the sender.

   o  The algorithm option is defined to carry the algorithms lists for
      algorithm agility.

   o  The signature is generated using the private key to verify the
      integrity of the DHCPv6 messages.  The Signature option is defined
      to carry the signature.

   o  The increasing number is used to detect replayed packet.  The
      Timestamp is one of the possible implementation choices.  The
BV> What is “The Timestamp”? Drop this sentence or perhaps provide more on what that is?




Jiang, et al.             Expires June 7, 2017                  [Page 6]

Internet-Draft                  SeDHCPv6                   December 2016


      Increasing-number option is defined to carry a strictly-increasing
      serial number.

   o  The encryption key Tag is calculated from the public key data.
      The Encryption Key Tag option is defined to identify the used
      public/private key pair.

   o  The Encrypted-message option is defined to contain the encrypted
      DHCPv6 message.

   o  The Encrypted-Query message is sent from the secure DHCPv6 client
      to the secure DHCPv6 server.  The Encrypted-Query message MUST
      contain the Encrypted-message option.  In addition, the Server
      Identifier option MUST be contained if it is contained in the
BV> chain be contained to be included? Kind of odd to so contained twice?
      original DHCPv6 message.  The Encrypted-Query message MUST NOT
      contain other options except the above options.
BV> Later on page 10 it is stated that the Encrypted-Query can also contain the Encryption-Key-Tag option?

   o  The Encrypted-Response message is sent from the secure DHCPv6
      server to the secure DHCPv6 client.  The Encrypted-Response
      message MUST contain the Encrypted-message option.  The Encrypted-
      Response message MUST NOT contain any other options except it.
BV> Drop “except it”? Other already makes that clear?


BV> So with respect to the Encrypted-Query and the Server Identifier option. When the client sends a Solicit (Confirm or Rebind) and there is no Server-Identifier, this will go to all servers but they probably won’t be able to decrypt? Does that make sense? Should the Server Identifier ALSO be in the Encrypted-Query (I’m NOT suggesting to allow it in the Client’s Solicit, Confirm, Rebind!)

BV> Also, given that the WG is working on Failover Protocol in parallel, how would seDHCPv6 work with Failover? And should we consider anything related to this. One possibility is that the failover partners could exchange some information (so we might have to define a new failover option eventually). But then my comment above about the Server Identifier may cause other issues? It may just be that once a client picks one of the failover partners, it will have to stick with that partner or return to the server discovery phase? But that would be too bad. (I had thought about whether the failover partners should share a Server Identifier, but that likely complicates other cases so not sure it is best.) Perhaps when the failover partners exchange the above mentioned information, that could include the Server Identifier which would allow them to be used.

5.3.  Support for Algorithm Agility

   In order to provide a means of addressing problems that may emerge
   with existing hash algorithms, signature algorithm and encryption
   algorithms in the future, this document provides a mechanism to
   support algorithm agility.  The support for algorithm agility in this
   document is mainly a algorithm notification mechanism between the
   client and the server.  The same client and server SHOULD use the
   same algorithm in a single communication session.  The sender can
   offer a set of algorithms, and then the receiver selects one
   algorithm for the future communication.

5.4.  Caused change to RFC3315

BV> Odd title. Perhaps “Impact on RFC3315” or something similar?

   For secure DHCPv6, the Solicit and Rebind messages can be sent only
   to the selected server(s) which share one common certificate.  If the
   client doesn't like the received Advertise(s) it could restart the
   whole process and selects another certificate, but it will be more
   expensive, and there's no guarantee that other servers can provide
   better Advertise(s).

   [RFC3315] provides an additional mechanism for preventing off-network
   timing attacks using the Reconfigure message: the Reconfigure Key
   authentication method.  Secure DHCPv6 can protect the Reconfigure
   message using the encryption method.  So the Reconfigure Key
   authentication method SHOULD NOT be used if Secure DHCPv6 is applied.



Jiang, et al.             Expires June 7, 2017                  [Page 7]

Internet-Draft                  SeDHCPv6                   December 2016


5.5.  Applicability

   In principle, secure DHCPv6 is applicable in any environment where
   physical security on the link is not assured and attacks on DHCPv6
   are a concern.  In practice, however, authenticated and encrypted
   DHCPv6 configuration will rely on some operational assumptions mainly
   regarding public key distribution and management.  In order to
   achieve the more wide use of secure DHCPv6, opportunistic security
BV> replace more wide with wider?
   [RFC7435] can be applied to secure DHCPv6 deployment, which allows
   DHCPv6 encryption in environments where support for authentication is
   not available.
BV> perhaps “for authentication or a key distribution mechanism”?

   Secure DHCPv6 can achieve authentication and encryption based on pre-
   sharing of authorized certificates.  The One feasible environment in
   an early deployment stage would be enterprise networks.  In
   enterprise networks, the client is manually pre-configured with the
   trusted servers' public key and the server is also manually pre-
   configured with the trusted clients' public keys.  In some scenario,
   such as coffee shop where the certificate cannot be validated and
   don't want to be blocked from the Internet, then the DHCPv6
BV> replace don’t want… with “one wants access to the Internet”?
   configuration process can be encrypted without authentication.

   Note that this deployment scenario based on manual operation is not
   different very much from the existing, shared-secret based
BV> replace different very much with “much different”?
   authentication mechanisms defined in [RFC3315] in terms of
   operational costs.  However, Secure DHCPv6 is still securer than the
   shared-secret mechanism in that even if clients' keys stored for the
   server are stolen that does not mean an immediate threat as these are
   public keys.  In addition, if some kind of PKI is used with Secure
BV> See earlier, but PKI not defined.
   DHCPv6, even if the initial installation of the certificates is done
   manually, it will help reduce operational costs of revocation in case
   a private key (especially that of the server) is compromised.

6.  DHCPv6 Client Behavior

   The secure DHCPv6 client is pre-configured with a certificate and its
   corresponding private key for client authentication.  If the client
   does not obtain a certificate from CA, it can generate the self-
BV> See earlier, CA not defined?
   signed certificate.

   The secure DHCPv6 client sends Information-request message as per
BV> sends an Information-request message (or sends Information-request messages)?
   [RFC3315].  The Information-request message is used by the DHCPv6
   client to request the server's certificate information without having
   addresses, prefixes or any non-security options assigned to it.  The
   contained Option Request option MUST carry the option code of the
   Certificate option.  In addition, the contained Algorithm option MUST
   be constructed as explained in Section 10.1.1.  The Information-




Jiang, et al.             Expires June 7, 2017                  [Page 8]

Internet-Draft                  SeDHCPv6                   December 2016


   request message MUST NOT include any other DHCPv6 options except the
   above options to minimize client's privacy information leakage.
BV> add the (the client’s)

   When receiving the Reply messages from DHCPv6 servers, a secure
BV> add the (the DHCPv6 servers)
   DHCPv6 client discards any DHCPv6 message that meets any of the
   following conditions:

   o  the Signature option is missing,

   o  multiple Signature options are present,

   o  the Certificate option is missing.

   And then the client first checks acknowledged hash, signature and
   encryption algorithms that the server supports.  If the hash
   algorithm field is zero, then it indicates that the hash algorithm is
   fixed according to the corresponding signature algorithm.  The client
   also uses the acknowledged algorithms in the return messages.

   Then the client checks the authority of the server.  The client
   validates the certificates through the pre-configured local trusted
   certificates list or other methods.  A certificate that finds a match
   in the local trust certificates list is treated as verified.  At this
   point, the client has either recognized the certificate of the
   server, or decided to drop the message.
BV> decided -> decide?

   The client MUST now authenticate the server by verifying the
   signature and checking increasing number, if there is a Increasing-
   number option.  The order of two procedures is left as an
BV> If this is the first message, what can client validate the Increasing Number against? From the flow, we are just in processing the Reply from the Information-Request so there is unlikely to be any information? Or is client expected to remember past values from a specific Server-ID?
   implementation decision.  It is RECOMMENDED to check increasing
   number first, because signature verification is much more
   computationally expensive.  The client checks the Increasing-number
   option according to the rule defined in Section 9.1 if it is
   contained.  For the message without an Increasing-number option,
   according to the client's local policy, it MAY be acceptable or
   rejected.  The Signature field verification MUST show that the
   signature has been calculated as specified in Section 10.1.3.  Only
   the messages that get through both the signature verification and
   increasing number check (if there is a Increasing-number option) are
   accepted.  Reply message that does not pass the above tests MUST be
   discarded.

   If there are multiple authenticated DHCPv6 certs, the client selects
BV> How are these multiple DHCPv6 certs received? I think you mean if multiple servers responded (hence multiple certs), the client has to pick one? I think stating this more clearly would be good?
   one DHCPv6 cert for the following communication.  The selected
   certificate may correspond to multiple DHCPv6 servers.  If there are
   no authenticated DHCPv6 certs or existing servers fail
   authentication, the client should retry a number of times.  The
   client conducts the server discovery process as per section 18.1.5 of



Jiang, et al.             Expires June 7, 2017                  [Page 9]

Internet-Draft                  SeDHCPv6                   December 2016


   [RFC3315] to avoid the packet storm.  In this way, it is difficult
BV> avoid a packet storm?
   for the rogue server to beat out a busy "real" server.  And then the
BV> change the rogue to a rogue?
   client takes some alternative action depending on its local policy,
   such as attempting to use an unsecured DHCPv6 server.

   Once the server has been authenticated, the DHCPv6 client sends the
   Encrypted-Query message to the DHCPv6 server.  The Encrypted-Query
   message contains the Encrypted-message option, which MUST be
   constructed as explained in Section 10.1.6.  The Encrypted-message
   option contains the encrypted DHCPv6 message using the public key
   contained in the selected cert.  In addition, the Server Identifier
   option MUST be included if it is in the original message (i.e.
   Request, Renew, Decline, Release) to avoid the need for other servers
   receiving the message to attempt to decrypt it.  The Encrypted-Query
   message MUST include the Encryption Key Tag option to identify the
   used public/private key pair, which is constructed as explained in
   Section 10.1.5.  The Encrypted-Query message MUST NOT contain any
BV> I don’t see that section 10.1.5 provides really any useful explanation as to how this data is constructed? Also, as I believe the server is supposed to use this information, how does it make sense for the client to generate it? If there is a clear algorithm that both the client and server can use to generate this (and the server has what it needs), OK. But more details are needed!!
   other DHCPv6 option except the Server Identifier option, Encryption
   Key Tag option, Encrypted-Message option.

   The first DHCPv6 message sent from the client to the server, such as
   Solicit message, MUST contain the Certificate option, Signature
   option and Increasing-number option for client authentication.  The
   encryption text SHOULD be formatted as explain in [RFC5652].  The
   Certificate option MUST be constructed as explained in
   Section 10.1.2.  In addition, one and only one Signature option MUST
   be contained, which MUST be constructed as explained in
  Section 10.1.3.  One and only one Increasing-number option SHOULD be
   contained, which MUST be constructed as explained in Section 10.1.4.
   In addition, the subsequent encrypted DHCPv6 message can also contain
   the Increasing-number option to defend against replay attack.
BV> Might help to clarify where this Increasing-Number option goes (in the client’s message).

   For the received Encrypted-Response message, the client MUST drop the
   Encrypted-Response message if other DHCPv6 option except Encrypted-
   message option is contained.  Then, the client extracts the
   Encrypted-message option and decrypts it using its private key to
   obtain the original DHCPv6 message.  In this document, it is assumed
   that the client uses only one certificate for the encrypted DHCPv6
   configuration.  So, the corresponding private key is used for
   decryption.  After the decryption, it handles the message as per
   [RFC3315].  If the decrypted DHCPv6 message contains the Increasing-
   number option, the DHCPv6 client checks it according to the rule
   defined in Section 9.1.

   If the client fails to get the proper parameters from the chosen
   server(s), it can select another authenticated certificate and send
   the Encrypted-Query message to another authenticated server(s) for




Jiang, et al.             Expires June 7, 2017                 [Page 10]

Internet-Draft                  SeDHCPv6                   December 2016


   parameters configuration until the client obtains the proper
   parameters.

   When the decrypted message is Reply message with an error status
   code, the error status code indicates the failure reason on the
   server side.  According to the received status code, the client MAY
   take follow-up action:

   o  Upon receiving an AuthenticationFail error status code, the client
      is not able to build up the secure communication with the server.
      However, there may be other DHCPv6 servers available that
      successfully complete authentication.  The client MAY use the
      AuthenticationFail as a hint and switch to other DHCPv6 server if
      it has another one.  The client SHOULD retry with another
      authenticated certificate.  However, if the client decides to
      retransmit using the same certificate after receiving
      AuthenticationFail, it MUST NOT retransmit immediately and MUST
      follow normal retransmission routines defined in [RFC3315].

   o  Upon receiving a DecryptionFail error status code, the client MAY
      resend the message following normal retransmission routines
      defined in [RFC3315].

   o  Upon receiving a ReplayDetected error status code, the client MAY
      resend the message with an adjusted Increasing-number option
      according to the returned number from the DHCPv6 server.

   o  Upon receiving a SignatureFail error status code, the client MAY
      resend the message following normal retransmission routines
      defined in [RFC3315].

BV> I’m really not sure whether these errors are helpful. Since how can the server even construct these in many cases and how can the client trust these if they aren’t authenticated?

(BV> I am still working through section 7 so may have more comments later). But, I did look at later sections so …


8.  Relay Agent Behavior

   When a DHCPv6 relay agent receives an Encrypted-query or Encrypted-
   response message, it may not recognize this message.  The unknown
   messages MUST be forwarded as described in [RFC7283].

   When a DHCPv6 relay agent recognizes the Encrypted-query and
   Encrypted-response messages, it forwards the message according to
   section 20 of [RFC3315].  There is nothing more the relay agents have
   to do, it neither needs to verify the messages from client or server,
   nor add any secure DHCPv6 options.  Actually, by definition in this
   document, relay agents MUST NOT add any secure DHCPv6 options.

   Relay-forward and Relay-reply messages MUST NOT contain any
   additional Certificate option or Increasing-number option, aside from
   those present in the innermost encapsulated messages from the client
   or server.

   Relay agent is RECOMMENDED to cache server announcements to form the
   list of the available DHCPv6 server certs.  If the relay agent
BV> NO! I do not think this is a wise idea at all. Please remove this! Relays should not do this!



Jiang, et al.             Expires June 7, 2017                 [Page 13]

Internet-Draft                  SeDHCPv6                   December 2016


   receives the Information-request message, then it replies with a list
   of server certs available locally.  In this way, the client can be
   confident of a quick response, and therefore treat the lack of a
   quick response as an indication that no authenticated DHCP servers
   exist.

9.  Processing Rules

9.1.  Increasing Number Check

   In order to check the Increasing-number option, defined in
   Section 10.1.4, the client/server has one stable stored number for
   replay attack detection.  The server should keep a record of the
   increasing number forever.  And the client keeps a record of the
   increasing number during the DHCPv6 configuration process with the
   DHCPv6 server.  And the client can forget the increasing number
   information after the transaction is finished.

   It is essential to remember that the increasing number is finite.
   All arithmetic dealing with sequence numbers must be performed modulo
   2^64.  This unsigned arithmetic preserves the relationship of
   sequence numbers as they cycle from 2^64 - 1 to 0 again.

   In order to check the Increasing-number option, the following
   comparison is needed.

   NUM.STO = the stored number in the client/server

   NUM.REC = the acknowledged number from the received message

   The Increasing-number option in the received message passes the
   increasing number check if NUM.REC is more than NUM.STO.  And then,
   the value of NUM.STO is changed into the value of NUM.REC.

   The increasing number check fails if NUM.REC is equal with or less
   than NUM.STO

10.  Extensions for Secure DHCPv6

   This section describes the extensions to DHCPv6.  Six new DHCPv6
   options, two new DHCPv6 messages and six new status codes are
   defined.

10.1.  New DHCPv6 Options







Jiang, et al.             Expires June 7, 2017                 [Page 14]

Internet-Draft                  SeDHCPv6                   December 2016


10.1.1.  Algorithm Option

   The Algorithm option carries the algorithms sets for algorithm
   agility, which is sent from the client to server.
BV> Might help if you indicated exactly where (in which message(s)) this option is in. Same goes for the other options defined later. In this case also, I think this is ONLY sent in the Information-Request message?

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |      OPTION_SIGNATURE         |         option-len            |
BV> Isn’t this OPTION_ALGORITHM
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   .                          EA-id List                           .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   .                          SA-id List                           .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   .                          HA-id List                           .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                        Figure 2: Algorithm Option

   o  option-code: OPTION_SIGNATURE (TBA1).

   o  option-len: length of EA-id List + length of SA-id List + length
      of HA-id List in octets.

   o  EA-id: The format of the EA-id List field is shown in Figure 3.

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |           EA-num              |               EA-id           |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   .                              ...                              .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |               EA-id           |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   EA-num         The number of the following EA-ids.
BV> Sadly RFC 7227 never defined a case like this and I think the WG should consider whether this should be LENGTH or a number of items. Personally, I prefer this to be a length (so EA-len which is 2 * number of EA-ids). Same goes for SA and HA. BUT I think this is something that the working group needs to consider. I was going to try to review other options to see if we had any that might resolve this but haven’t do that review (though nothing obvious comes to mind).

   EA-id          Encryption Algorithm id. The encryption algorithm
                  is used for the encrypted DHCPv6 configuration
                  process. This design is adopted in order to provide
                  encryption algorithm agility. The value is from the
                  Encryption Algorithm for Secure DHCPv6 registry in
                  IANA. A registry of the initial assigned values
                  is defined in Section 12. The mandatory encryption
                  algorithms MUST be included.
BV> This never says that this is a 2-octet value? Please be clear about that. Same goes for SA and HA below.
BV> I think also saying something like “The client enumerates the list of Encryption algorithms it supports to the server” would be extremely useful for the EA-List (with similar text for the SA and HA lists)?

                        Figure 3: EA-id List Field



Jiang, et al.             Expires June 7, 2017                 [Page 15]

Internet-Draft                  SeDHCPv6                   December 2016


   o  SA-id List: The format of the SA-id List field is shown in
      Figure 4.

0                   1                   2                   3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|           SA-num              |               SA-id           |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
.                              ...                              .
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|               SA-id           |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

SA-num         The number of the following SA-ids.

SA-id          Signature Algorithm id. This design is adopted in
               order to provide signature algorithm agility. The
               value is from the Signature Algorithm for Secure
               DHCPv6 registry in IANA. The support of RSASSA-PKCS1-v1_5
               is mandatory. A registry of the initial assigned
               values is defined in Section 12. The mandatory
               signature algorithms MUST be included.

                        Figure 4: SA-id List Field

   o  HA-id List: The format of the HA-id List field is shown in
      Figure 5.

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |           HA-num              |               HA-id           |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   .                              ...                              .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |               HA-id           |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   HA-num         The number of the following HA-ids.

   HA-id          Hash Algorithm id. This design is adopted in order to
                  provide hash algorithm agility. The value is from the
                  Hash Algorithm for Secure DHCPv6 registry in IANA. The
                  support of SHA-256 is mandatory. A registry of the
                  initial assigned values is defined in Section 12.
                  The mandatory hash algorithms MUST be included.

                        Figure 5: HA-id List Field



Jiang, et al.             Expires June 7, 2017                 [Page 16]

Internet-Draft                  SeDHCPv6                   December 2016


10.1.2.  Certificate Option

   The Certificate option carries the certificate of the client/server.
   The format of the Certificate option is described as follows:

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |      OPTION_CERTIFICATE       |         option-len            |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |              EA-id            |            SA-id              |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |                                                               |
   .                           Certificate                         .
   |                                                               |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                       Figure 6: Certificate Option

   o  option-code: OPTION_CERTIFICATE (TBA2).

   o  option-len: 4 + length of Certificate in octets.

   o  EA-id: Encryption Algorithm id.  The encryption algorithm is used
      for the encrypted DHCPv6 configuration process.  This design is
      adopted in order to provide encryption algorithm agility.  The
      value is from the Encryption Algorithm for Secure DHCPv6 registry
      in IANA.  A registry of the initial assigned values is defined in
      Section 12.  If the value of EA-id is 0, then the certificate is
      not used for encryption.
BV> This seems a bit odd … shouldn’t this just say the Encryption Algorithm ID used for this certificate?

   o  SA-id: Signature Algorithm id.  The signature algorithm is used
      for computing the signature result.  The value is from the
      Signature Algorithm for Secure DHCPv6 registry in IANA.  A
      registry of the initial assigned values is defined in Section 12.
      If the value of SA-id is 0, then the certificate is not used for
      signature check.
BV> Same – see above for EA-id?

   o  Certificate: A variable-length field containing certificates.  The
      encoding of certificate and certificate data MUST be in format as
      defined in Section 3.6, [RFC7296].  The support of X.509
      certificate is mandatory.

   It should be noticed that the scenario where the values of EA-id and
   SA-id are all 0, it makes no sense and MUST NOT be used.
BV> Perhaps say “… are both 0 makes no sense and the client MUST discard a message with such values”?






Jiang, et al.             Expires June 7, 2017                 [Page 17]

Internet-Draft                  SeDHCPv6                   December 2016


10.1.3.  Signature option

   The Signature option allows a signature that is signed by the private
   key to be attached to a DHCPv6 message.  The Signature option could
   be in any place within the DHCPv6 message while it is logically
   created after the entire DHCPv6 header and options.  It protects the
   entire DHCPv6 header and options, including itself.  The format of
   the Signature option is described as follows:

    0                   1                   2                   3
   0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |     OPTION_SIGNATURE          |        option-len             |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |         SA-id                 |            HA-id              |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |                                                               |
   .                    Signature (variable length)                .
   .                                                               .
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                        Figure 7: Signature Option

   o  option-code: OPTION_SIGNATURE (TBA3).

   o  option-len: 4 + length of Signature field in octets.

   o  SA-id: Signature Algorithm id.  The signature algorithm is used
      for computing the signature result.  This design is adopted in
      order to provide signature algorithm agility.  The value is from
      the Signature Algorithm for Secure DHCPv6 registry in IANA.  The
      support of RSASSA-PKCS1-v1_5 is mandatory.  A registry of the
      initial assigned values is defined in Section 12.

   o  HA-id: Hash Algorithm id.  The hash algorithm is used for
      computing the signature result.  This design is adopted in order
      to provide hash algorithm agility.  The value is from the Hash
      Algorithm for Secure DHCPv6 registry in IANA.  The support of
      SHA-256 is mandatory.  A registry of the initial assigned values
      is defined in Section 12.  If the hash algorithm is fixed
      according to the corresponding signature algorithm, the HA-id
      field is set to zero.

   o  Signature: A variable-length field containing a digital signature.
      The signature value is computed with the hash algorithm and the
      signature algorithm, as described in HA-id and SA-id.  The
      Signature field MUST be padded, with all 0, to the next octet
      boundary if its size is not a multiple of 8 bits.  The padding



Jiang, et al.             Expires June 7, 2017                 [Page 18]

Internet-Draft                  SeDHCPv6                   December 2016


      length depends on the signature algorithm, which is indicated in
      the SA-id field.

   Note: If Secure DHCPv6 is used, the DHCPv6 message is encrypted in a
   way that the authentication mechanism defined in RFC3315 does not
   understand.  So the Authentication option SHOULD NOT be used if
   Secure DHCPv6 is applied.

10.1.4.  Increasing-number Option

   The Increasing-number option carries the strictly increasing number
   for anti-replay protection.  It is optional.

0                   1                   2                   3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|   OPTION_INCREASING_NUM       |        option-len             |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|                                                               |
|                  InreasingNum (64-bit)                        |
BV> Increasing-Num? (also use – below?)
|                                                               |
|                                                               |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

option-code    OPTION_INCREASING_NUM (TBA4).

option-len     8, in octets.

IncreasingNum  A strictly increasing number for the replay attack detection
               which is more than the local stored number.

                    Figure 8: Increasing-number Option

10.1.5.  Encryption Key Tag Option

   The Encryption Key Tag option carries the key identifier which is
   calculated from the public key data.  The Encrypted-Query message
   MUST contain the Encryption Key Tag option to identify the used
   public/private key pair.












Jiang, et al.             Expires June 7, 2017                 [Page 19]

Internet-Draft                  SeDHCPv6                   December 2016


      0                   1                   2                   3
      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |          option-code          |           option-len          |
BV> Earlier you placed option name – OPTION_ENCRYPTION_KEY_TAG
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     .                    encryption key tag                         .
     .                       (variable)                              .
     .                                                               .
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                    Figure 9: Encryption Key Tag Option

   option-code  OPTION_ENCRY_KT (TBA5).
BV> I’d recommend OPTION_ENCRYPTION_KEY_TAG as it isn’t that long and more clear?

   option-len  Length of the encryption key tag.

   encryption key tag  A variable length field containing the encryption
      key tag sent from the client to server to identify the used
      public/private key pair.  The encryption key tag is calculated
      from the public key data, like fingerprint of a specific public
      key.

BV> This is really unclear to me? How is this calculated? I think more details are needed as client generates this from server’s data so server must know how client generated it?

10.1.6.  Encrypted-message Option

   The Encrypted-message option carries the encrypted DHCPv6 message,
   which is calculated with the recipient's public key.

   The format of the Encrypted-message option is:

      0                   1                   2                   3
      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |          option-code          |           option-len          |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     .                  encrypted DHCPv6 message                     .
     .                       (variable)                              .
     .                                                               .
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                    Figure 10: Encrypted-message Option

   option-code  OPTION_ENCRYPTED_MSG (TBA6).

   option-len  Length of the encrypted DHCPv6 message.





Jiang, et al.             Expires June 7, 2017                 [Page 20]

Internet-Draft                  SeDHCPv6                   December 2016


   encrypted DHCPv6 message  A variable length field containing the
      encrypted DHCPv6 message.  In Encrypted-Query message, it contains
      encrypted DHCPv6 message sent from a client to server.  In
      Encrypted-response message, it contains encrypted DHCPv6 message
      sent from a server to client.

10.2.  New DHCPv6 Messages

   Two new DHCPv6 messages are defined to achieve the DHCPv6 encryption:
   Encrypted-Query and Encrypted-Response.  Both the DHCPv6 messages
   defined in this document share the following format:

      0                   1                   2                   3
      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |    msg-type   |               transaction-id                  |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     .                             options                           .
     .                           (variable)                          .
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

      Figure 11: The format of Encrypted-Query and Encrypted-Response
                                 Messages

   msg-type        Identifier of the message type.  It can be either
                   Encrypted-Query (TBA7) or DHCPv6-Response (TBA8).

   transaction-id  The transaction ID for this message exchange.

   options         The Encrypted-Query message MUST contain the
                   Encrypted-message option, Encryption Key Tag option
                   and Server Identifier option if the message in the
                   Encrypted-message option has a Server Identifier
                   option.  The Encrypted-Response message MUST only
                   contain the Encrypted-message option.

10.3.  Status Codes

   The following new status codes, see Section 5.4 of [RFC3315] are
   defined.

   o  AuthenticationFail (TBD9): indicates that the message from the
      DHCPv6 client fails authentication check.

   o  ReplayDetected (TBD10): indicates the message from DHCPv6 client
      fails the increasing number check.



Jiang, et al.             Expires June 7, 2017                 [Page 21]

Internet-Draft                  SeDHCPv6                   December 2016


   o  SignatureFail (TBD11): indicates the message from DHCPv6 client
      fails the signature check.

11.  Security Considerations

   This document provides the authentication and encryption mechanisms
   for DHCPv6.

   [RFC6273] has analyzed possible threats to the hash algorithms used
   in SEND.  Since Secure DHCPv6 defined in this document uses the same
   hash algorithms in similar way to SEND, analysis results could be
   applied as well: current attacks on hash functions do not constitute
   any practical threat to the digital signatures used in the signature
   algorithm in Secure DHCPv6.

   A server, whose local policy accepts messages without a Increasing-
   number option, may have to face the risk of replay attacks.
BV> Isn’t this also an issue for the client?

   There are some mandatory algorithm for encryption algorithm in this
   document.  It may be at some point that the mandatory algorithm is no
   longer safe to use.
BV> Isn’t this related to the 2nd paragraph in this section? Perhaps move this as third paragraph?

   If the client tries more than one cert for client authentication, the
   server can easily get a client that implements this to enumerate its
   entire cert list and probably learn a lot about a client that way.
BV> Are there any recommendations we could make about this? Perhaps this is more related to client configuration but perhaps in some cases client certificates could be tied to specific server certificates by configuration?


-          Bernie

From: dhcwg [mailto:dhcwg-bounces@ietf.org] On Behalf Of Lishan Li
Sent: Monday, December 05, 2016 10:59 AM
To: dhcwg <dhcwg@ietf.org>;
Subject: Re: [dhcwg] I-D Action: draft-ietf-dhc-sedhcpv6-18.txt

Dear All,

We have submitted the new version of secure DHCPv6.
In the new version, we made the following updates:
1. Add the Algorithm option, Encryption Key Tag option;
2. Delete the AlgorithmNotSupported error status code;
3. Delete some description on that secure DHCPv6
    exchanges the server selection method;
4. Add the assumption that: For DHCPv6 client, just one
    certificate is used for the DHCPv6 configuration;
5. Add the statement that: For the first Encrypted-Query
    message, the server needs to try all the possible private
    keys and then records the relationship between the public
    key and the encryption key tag;
5. For the case where the client's certificate is missed
    and decryption fails, the server discards the received
    message.

Could you please review the current version? Looking
forward to your guidance.

Best Regards,
Lishan

2016-12-05 16:03 GMT+08:00 <internet-drafts@ietf.org<mailto:internet-drafts@ietf.org>>:

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the Dynamic Host Configuration of the IETF.

        Title           : Secure DHCPv6
        Authors         : Sheng Jiang
                          Lishan Li
                          Yong Cui
                          Tatuya Jinmei
                          Ted Lemon
                          Dacheng Zhang
        Filename        : draft-ietf-dhc-sedhcpv6-18.txt
        Pages           : 30
        Date            : 2016-12-05

Abstract:
   DHCPv6 includes no deployable security mechanism that can protect
   end-to-end communication between DHCP clients and servers.  This
   document describes a mechanism for using public key cryptography to
   provide such security.  The mechanism provides encryption in all
   cases, and can be used for authentication based on pre-sharing of
   authorized certificates.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-dhc-sedhcpv6/

There's also a htmlized version available at:
https://tools.ietf.org/html/draft-ietf-dhc-sedhcpv6-18

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-dhc-sedhcpv6-18


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org<http://tools.ietf.org>.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org<mailto:dhcwg@ietf.org>
https://www.ietf.org/mailman/listinfo/dhcwg