[Docs] [txt|pdf] [Tracker] [Email] [Diff1] [Diff2] [Nits]

Versions: 00 01 02

Network Working Group                                            F. Gont
Internet-Draft                                    SI6 Networks / UTN-FRH
Intended status: Informational                                 R. Bonica
Expires: April 25, 2014                                 Juniper Networks
                                                                  W. Liu
                                                     Huawei Technologies
                                                        October 22, 2013


        Security Assessment of Neighbor Discovery (ND) for IPv6
                  draft-gont-opsec-ipv6-nd-security-02

Abstract

   Neighbor Discovery is one of the core protocols of the IPv6 suite,
   and provides in IPv6 similar functions to those provided in the IPv4
   protocol suite by the Address Resolution Protocol (ARP) and the
   Internet Control Message Protocol (ICMP).  Its increased flexibility
   implies a somewhat increased complexity, which has resulted in a
   number of bugs and vulnerabilities found in popular implementations.
   This document provides guidance in the implementation of Neighbor
   Discovery, and documents issues that have affected popular
   implementations, in the hopes that the same issues do not repeat in
   other implementations.

Status of this Memo

   This Internet-Draft is submitted in full conformance with the
   provisions of BCP 78 and BCP 79.  This document may not be modified,
   and derivative works of it may not be created, and it may not be
   published except as an Internet-Draft.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF).  Note that other groups may also distribute
   working documents as Internet-Drafts.  The list of current Internet-
   Drafts is at http://datatracker.ietf.org/drafts/current/.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   This Internet-Draft will expire on April 25, 2014.

Copyright Notice

   Copyright (c) 2013 IETF Trust and the persons identified as the
   document authors.  All rights reserved.



Gont, et al.             Expires April 25, 2014                 [Page 1]

Internet-Draft           ND Security Assessment             October 2013


   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.


Table of Contents

   1.  DISCLAIMER . . . . . . . . . . . . . . . . . . . . . . . . . .  4
   2.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  5
   3.  Neighbor Discovery messages  . . . . . . . . . . . . . . . . .  6
     3.1.  Router Solicitation message  . . . . . . . . . . . . . . .  6
     3.2.  Router Advertisement . . . . . . . . . . . . . . . . . . .  7
     3.3.  Neighbor Solicitation message  . . . . . . . . . . . . . . 11
     3.4.  Neighbor Advertisement message . . . . . . . . . . . . . . 12
     3.5.  Redirect message . . . . . . . . . . . . . . . . . . . . . 15
     3.6.  Neighbor Discovery Options . . . . . . . . . . . . . . . . 18
       3.6.1.  General issues with Neighbor Discovery options . . . . 19
       3.6.2.  Source Link-Layer Address Option . . . . . . . . . . . 20
       3.6.3.  Target Link-Layer Address Option . . . . . . . . . . . 22
       3.6.4.  Prefix Information . . . . . . . . . . . . . . . . . . 23
       3.6.5.  Redirected Header Option . . . . . . . . . . . . . . . 26
       3.6.6.  MTU Option . . . . . . . . . . . . . . . . . . . . . . 27
       3.6.7.  Route Information Option . . . . . . . . . . . . . . . 28
       3.6.8.  Recursive DNS Server Option  . . . . . . . . . . . . . 31
       3.6.9.  DNS Search List  . . . . . . . . . . . . . . . . . . . 33
   4.  Router and Prefix Discovery  . . . . . . . . . . . . . . . . . 34
     4.1.  Router Specification . . . . . . . . . . . . . . . . . . . 34
     4.2.  Host Specification . . . . . . . . . . . . . . . . . . . . 34
   5.  Address Resolution . . . . . . . . . . . . . . . . . . . . . . 36
     5.1.  Interface initialization . . . . . . . . . . . . . . . . . 38
     5.2.  Receipt of Neighbor Solicitation messages  . . . . . . . . 39
   6.  Vulnerability analysis . . . . . . . . . . . . . . . . . . . . 40
     6.1.  Denial of Service  . . . . . . . . . . . . . . . . . . . . 40
       6.1.1.  Neighbor Cache poisoning . . . . . . . . . . . . . . . 41
       6.1.2.  Tampering with Duplicate Address Detection (DAD) . . . 41
       6.1.3.  Tampering with Neighbor Unreachability Detection
               (NUD)  . . . . . . . . . . . . . . . . . . . . . . . . 42
       6.1.4.  Rogue Router . . . . . . . . . . . . . . . . . . . . . 43
       6.1.5.  Parameter spoofing . . . . . . . . . . . . . . . . . . 43
       6.1.6.  Bogus on-link prefixes . . . . . . . . . . . . . . . . 44
       6.1.7.  Bogus address configuration prefixes . . . . . . . . . 45
       6.1.8.  Disabling routers  . . . . . . . . . . . . . . . . . . 45



Gont, et al.             Expires April 25, 2014                 [Page 2]

Internet-Draft           ND Security Assessment             October 2013


       6.1.9.  Tampering with 'on-link determination' . . . . . . . . 46
       6.1.10. Introducing forwarding loops at routers  . . . . . . . 48
       6.1.11. Tampering with a Neighbor Discovery implementation . . 49
       6.1.12. Tampering with a Neighbor Discovery router
               implementation from a remote site  . . . . . . . . . . 51
     6.2.  Performance degrading  . . . . . . . . . . . . . . . . . . 52
       6.2.1.  Parameter spoofing . . . . . . . . . . . . . . . . . . 52
     6.3.  Traffic hijacking  . . . . . . . . . . . . . . . . . . . . 52
       6.3.1.  Neighbor Cache poisoning . . . . . . . . . . . . . . . 52
       6.3.2.  Rogue Router . . . . . . . . . . . . . . . . . . . . . 53
       6.3.3.  Bogus on-link prefixes . . . . . . . . . . . . . . . . 53
       6.3.4.  Tampering with 'on-link determination' . . . . . . . . 54
     6.4.  Miscellaneous security issues  . . . . . . . . . . . . . . 54
       6.4.1.  Detecting Sniffing Hosts . . . . . . . . . . . . . . . 54
   7.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 55
   8.  Security Considerations  . . . . . . . . . . . . . . . . . . . 56
   9.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 57
   10. References . . . . . . . . . . . . . . . . . . . . . . . . . . 58
     10.1. Normative References . . . . . . . . . . . . . . . . . . . 58
     10.2. Informative References . . . . . . . . . . . . . . . . . . 59
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 62






























Gont, et al.             Expires April 25, 2014                 [Page 3]

Internet-Draft           ND Security Assessment             October 2013


1.  DISCLAIMER

   This is WORK IN PROGRESS.  Some of the recommendations might possibly
   change.  For instance, some (NOT all) of the proposed "sanity checks"
   help reduce vulnerability to some attacks at the expense of e.g.
   reduced responsiveness.  Further discussion might find some of such
   checks to be inadequate or inappropriate.  On the other hand, some of
   mitigations discussed in this document have been incorporated into
   popular Neighbor Discovery (ND) implementations.










































Gont, et al.             Expires April 25, 2014                 [Page 4]

Internet-Draft           ND Security Assessment             October 2013


2.  Introduction

   Neighbor Discovery is used by nodes on the same link to discover each
   other's presence, to determine each other's link-layer addresses, to
   find routers, and to maintain reachability information about the
   paths to active neighbors [RFC4861].

   Neighbor Discovery is specified by [RFC4861].  [RFC3122] specifies
   extensions to Neighbor Discovery for Inverse Discovery.  [RFC4389]
   specifies Neighbor Discovery proxies.  [RFC3756] describes trust
   models and threats for Neighbor Discovery.  [RFC3971] specifies a
   secure version of Neighbor Discovery named 'SEcure Neighbor Discovery
   (SEND)'.

   Neighbor Discovery was originally specified by [RFC2461], which was
   later obsoleted by [RFC4861].  [RFC4943] clarifies the rationale for
   the removal of the 'on-link assumption' from [RFC4861].

   Section 3 of this document provides an analysis of each of the
   Neighbor Discovery messages, along with a discussion of the Neighbor
   Discovery options that have been specified at the time of this
   writing.  Section 4 discusses the security implications of Router and
   Prefix Discovery.  Section 5 describes the security implications of
   Address Resolution.  Section 6 contains a vulnerability analysis of
   Neighbor Discovery.

   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 RFC 2119 [RFC2119].






















Gont, et al.             Expires April 25, 2014                 [Page 5]

Internet-Draft           ND Security Assessment             October 2013


3.  Neighbor Discovery messages

   The following subsections discuss a number of validation checks that
   should be performed on Neighbor Discovery messages.

3.1.  Router Solicitation message

   The following figure illustrates the syntax of Router Solicitation
   messages:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Code      |          Checksum             |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                            Reserved                           |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |   Options ...
     +-+-+-+-+-+-+-+-+-+-+-+-

            Figure 1: ICMPv6 Router Solicitation message format

   As can be inferred from syntax of Router Solicitation messages, any
   legitimate Router Solicitation message must have a length (as derived
   from the IPv6 length) that is 8 octets or more.  If the packet does
   not pass this check, it should be silently dropped.

   The Source Address of an IPv6 packet encapsulating a Router
   Solicitation message is set to the value of one of the addresses
   assigned to the sending interface, or to the unspecified address (::)
   if no address has been assigned to that interface.  Nodes should
   discard Router Solicitation messages that have a multicast address in
   the Source Address field.

   The Destination Address of an IPv6 packet encapsulating a Router
   Solicitation message is set to the all-routers multicast address.

      A unicast address could possibly be used for the Destination
      Address for debugging purposes.

   If a unicast address is used for the Destination Address, the
   receiving system should ensure that it is a link-local address.  If
   the packet does not pass this check, it should be silently dropped.







Gont, et al.             Expires April 25, 2014                 [Page 6]

Internet-Draft           ND Security Assessment             October 2013


      While this is not explicitly required in [RFC4861] this provides
      an additional counter-measure (other than the validation of the
      Hop Limit) for non-local malicious nodes willing to make use of
      Router Solicitation messages for reconnaissance purposes.

   As of this writing, the following options are valid in a Router
   Solicitation message:

   o  Source link-layer address

   Any other options should be silently ignored.

   If a 'source link-layer address' option is included, the following
   sanity checks should be performed:

   o  The Source Address of the packet must not the unspecified address
      (::) or the "loopback" addresses (::1)

   o  The advertised link-layer address must not a broadcast or
      multicast address

3.2.  Router Advertisement

   The following figure illustrates the syntax of Router Advertisement
   messages.


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Code      |          Checksum             |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     | Cur Hop Limit |M|O|H|Prf|Resvd|       Router Lifetime         |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                         Reachable Time                        |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                          Retrans Timer                        |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |   Options ...
     +-+-+-+-+-+-+-+-+-+-+-+-

           Figure 2: ICMPv6 Router Advertisement message format

   The Source Address of an IPv6 packet encapsulating a Router
   Advertisement message is set to a link-local address assigned to the
   interface from which the message is sent.  Nodes should discard
   Router Advertisements whose Source Address is not a link-local
   address.



Gont, et al.             Expires April 25, 2014                 [Page 7]

Internet-Draft           ND Security Assessment             October 2013


   The Destination Address of an IPv6 packet encapsulating a Router
   Advertisement message is set to the Source Address of the system that
   elicited the Router Advertisement message (unless this was the
   unspecified address), or in the case of unsolicited Router
   Advertisements, to the all-nodes multicast address.  Nodes receiving
   a Router Advertisement should ensure that if the Destination Address
   is a unicast address, it is a link-local address.  Otherwise, the
   Router Advertisement message should be silently dropped.

      While this is not explicitly required in [RFC4861] this provides
      another mitigation for non-local malicious nodes willing to make
      use of Router Solicitation messages for reconnaissance purposes.

   The Cur Hop Limit field specifies the default value that should be
   placed in the Hop Count field of outgoing IPv6 packets.  As stated in
   [RFC4861] a value of 0 means unspecified (by this router).  If the
   Cur Hop Limit field is larger than 0, nodes should sanitize the
   received Cur Hop Limit value as follows:

              SanitizedCH = max(Cur Hop Limit, MIN_HOP_LIMIT)

   where the sanitized Cur Hop Limit (SanitizedCH) is set to the maximum
   of the Cur Hop Limit and the variable MIN_HOP_LIMIT.  MIN_HOP_LIMIT
   should default to 64, and should be configurable by the system
   administrator.

   If the received Cur Hop Limit were not sanitized, an attacker could
   perform a Denial-of-Service (DoS) attack against the local network by
   forging a Router Advertisement message that includes a very small Cur
   Hop Limit value.  As a result, nodes honouring the Router
   Advertisement would set the Hop Limit of outgoing packets to such
   small value, and as a result those packets would be dropped by some
   intervening router.

   For example, if an attacker were to forge a Router Advertisement that
   contains a Cur Hop Limit of 1, the victim nodes could communicate
   only with nodes on the same network link, as their packets would be
   dropped by the first-hop router.

   XXXX The Prf field is specified in [RFC4191] and is used to specify a
   'preference' value for the router sending the Router Advertisement.

   The Router Lifetime field is a 16-bit unsigned integer that specifies
   the lifetime associated with the default router in units of seconds.
   A Router Lifetime of 0 indicates that the router is not a default
   router and must not appear in the default router list.  The sending
   rules in Section 6 of [RFC4861] limit the Router Lifetime to 9000
   seconds.  However, nodes are expected to handle any value.



Gont, et al.             Expires April 25, 2014                 [Page 8]

Internet-Draft           ND Security Assessment             October 2013


      An attacker could exploit the Router Lifetime field to perform DoS
      attacks or performance-degrading attacks.  For example, an
      attacker could forge Router Advertisement messages that include a
      very small Router Lifetime.  This would have a two-fold effect on
      the network.  Firstly, once the advertised router expires as a
      'default' router, the corresponding nodes might face a Denial of
      Service, as a result of having no default routers.  Secondly, a
      small Router Lifetime value could lead to increased traffic in the
      network, and increased processing time in the affected nodes (as a
      result of the additional Router Solicitation/Advertisement
      exchanges needed to re-configure the routing table of each node).

   If the Router Lifetime is different from 0, it should be sanitized as
   follows:

       SanitizedRL = min( max(Router Lifetime, MIN_ROUTER_LIFETIME),
                          MAX_ROUTER_LIFETIME)

   where lower and upper limits are enforced on the advertised Router
   Lifetime.  The lower limit is specified by the variable
   MIN_ROUTER_LIFETIME, and should default to 1800 seconds.  The upper
   limit is specified by MAX_ROUTER_LIFETIME, and should default to 9000
   seconds.

   The value '1800 seconds' results from the recommended default value
   (AdvDefaultLifetime) for setting the Router Lifetime, which instead
   results from the expression '3 * MaxRtrAdvInterval' (where
   MaxRtrAdvInterval defaults to 600 seconds).  The value '9000 seconds'
   results from the required upper limit for setting the Router Lifetime
   field (AdvDefaultLifetime).

   The Router Lifetime should not be sanitized when it is equal to 0, as
   a value of 0 indicates that the corresponding router should not be
   used as a default router (i.e., it is only advertising prefixes).

   When a router is in the Default routers list, and a Router
   Advertisement is received with a Router Lifetime of 0, a node might
   choose to keep the router in the Default routers list (as allowed by
   the current local Router Lifetime value).  This might allow nodes to
   be resilient to Router Advertisements that incorrectly or maliciously
   advertise a Router Lifetime of 0, at the expense of loss of
   responsiveness in scenarios in which a router explicitly advertises
   it wants to be removed from the Default routers list (such a scenario
   is described in Section 6.2.5 of [RFC4861].

   The Reachable Time field is a 32-bit unsigned integer that specifies
   the amount of time, in milliseconds, that a node assumes a neighbor
   is reachable after having received a reachability confirmation.  A



Gont, et al.             Expires April 25, 2014                 [Page 9]

Internet-Draft           ND Security Assessment             October 2013


   value of zero means 'unspecified by this router'.  If Reachable Time
   is different from 0, it should be sanitized as follows:

        SanitizedRT = max( min( Reachable Time, MAX_REACHABLE_TIME),
                           MIN_REACHABLE_TIME)

   where MAX_REACHABLE_TIME and MIN_REACHABLE_TIME impose upper and
   lower limits, respectively, to the received Reachable Time value.  We
   propose a MAX_REACHABLE_TIME of 3,600,000 (one hour) and a
   MIN_REACHABLE_TIME of 20,000.

   The upper limit of 3,600,000 is specified in Section 6.2.1 of
   [RFC4861] (AdvReachableTime router variable).  The lower limit has
   been selected such that the minimum local ReachableTime (that would
   result from MIN_RANDOM_FACTOR * SanitizedRT) is not smaller than 10
   seconds.

   The Retrans Timer is a 32-bit unsigned integer that specifies the
   amount of time, in milliseconds between retransmitted Neighbor
   Solicitation messages.  A value of zero means 'unspecified by this
   router'.  If Retrans Timer is different from 0, it should be
   sanitized as follows:

         SanitizedRXT = max( min( Retrans Timer, MAX_RETRANS_TIME),
                             MIN_RETRANS_TIME)

   We propose a MAX_RETRANS_TIME of 60,000 and a MIN_RETRANS_TIME of
   1,000.

   At the time of this writing, the options that may be legitimately
   included in Router Advertisements are:

   o  Source link-layer address

   o  MTU

   o  Prefix information

   o  Route Information

   o  Recursive DNS Server

   o  DNS Search List

   Other options should be silently ignored.

   The Source link-layer address option specifies the link-layer address
   of the interface from which the Router Advertisement is sent.  It is



Gont, et al.             Expires April 25, 2014                [Page 10]

Internet-Draft           ND Security Assessment             October 2013


   only used on link layers that have addresses.  Nodes should ignore
   the source link-layer address option in Router Advertisements
   received on link layers that do not have addresses.

   Section 3.6 of this document discusses the security implications of
   all the Neighbor Discovery options.

3.3.  Neighbor Solicitation message

   The following figure illustrates the format of Neighbor Solicitation
   messages:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Code      |          Checksum             |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Reserved                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     +                                                               +
     |                                                               |
     +                       Target Address                          +
     |                                                               |
     +                                                               +
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |   Options ...
     +-+-+-+-+-+-+-+-+-+-+-+-

           Figure 3: ICMPv6 Neighbor Solicitation message format

   The Source Address of an IPv6 packet encapsulating a Neighbor
   Solicitation message is set to an address assigned to the interface
   from which the message is sent, or to the unspecified address (::).

   The Destination Address of an IPv6 packet encapsulating a Neighbor
   Solicitation message is set to the solicited-node multicast address
   corresponding to the target address, or to the target address.

   The ICMPv6 packet length (as derived from the IPv6 Payload Length)
   must be greater than or equal to 24.  If the packet does not pass
   this check, it should be silently dropped.

   The Target Address is the IPv6 address of the target of the
   solicitation.  The Target Address must pass the following checks:




Gont, et al.             Expires April 25, 2014                [Page 11]

Internet-Draft           ND Security Assessment             October 2013


   1.  It must not be a multicast address (as required in Section 4.3 of
       [RFC4861])

   2.  It must not be the unspecified address (::)

   3.  It must not be the loopback address (::1)

   The Target Address must also meet any of the following criteria:

   1.  It is a valid unicast or anycast address assigned to the
       receiving interface

   2.  It is a unicast or anycast address for which the node is offering
       proxy service

   3.  It is a 'tentative' address on which 'Duplicate Address
       Detection' (DAD) is being performed (in which case the Neighbor
       Solicitation message should be processed according to [RFC4862])

   At the time of this writing, the options that may be legitimately
   included in Neighbor Solicitations are:

   o  Source link-layer address

   According to Section 4.3 of [RFC4861], the source link-layer address
   option must not be included when the Source Address is the
   unspecified address (::).  A node receiving a Neighbor Solicitation
   that includes a source link-layer address and that has the
   unspecified address (::) as the Source Address should silently drop
   the corresponding packet.

   According to Section 4.3 of [RFC4861], on link layers that have
   addresses (and provided that the Source Address is not the
   unspecified address), Neighbor Solicitations sent to multicast
   addresses must include the source link-layer address option.  A node
   receiving a Neighbor Solicitation sent to a multicast address that
   does not include a source link-layer option should be silently
   dropped.

3.4.  Neighbor Advertisement message

   A node sends Neighbor Advertisements in response to Neighbor
   Solicitations and sends unsolicited Neighbor Advertisements in order
   to (unreliably) propagate new information quickly [RFC4861].

   The following figure illustrates the syntax of Neighbor Advertisement
   messages:




Gont, et al.             Expires April 25, 2014                [Page 12]

Internet-Draft           ND Security Assessment             October 2013


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Code      |          Checksum             |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |R|S|O|                     Reserved                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     +                                                               +
     |                                                               |
     +                       Target Address                          +
     |                                                               |
     +                                                               +
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |   Options ...
     +-+-+-+-+-+-+-+-+-+-+-+-

          Figure 4: ICMPv6 Neighbor Advertisement message format

   The Source Address of an IPv6 packet encapsulating a Neighbor
   Advertisement message is set to a link-local address assigned to the
   interface from which the message is sent.  Nodes should discard
   Neighbor Advertisements that do not have a link-local address in the
   Source Address field.

   The Destination Address of an IPv6 packet encapsulating a Neighbor
   Advertisement message is set to the Source Address of the Neighbor
   Solicitation that elicited the Neighbor Advertisement message
   (provided the Source Address of the Neighbor Solicitation was a
   unicast address).  If the Source Address of the Neighbor Solicitation
   was the unspecified address, the Neighbor Advertisement is sent to
   the all-nodes multicast address.  Finally, unsolicited Neighbor
   Advertisements are sent to the all-nodes multicast address

   The Hop Limit of an IPv6 packet encapsulating a Neighbor
   Advertisement message must be set to 255 by the sending node.  A node
   receiving a Neighbor Advertisement message should perform the
   following check:

   The ICMPv6 packet length (as derived from the IPv6 Payload Length)
   must be greater than or equal to 24.  If the packet does not pass
   this check, it should be silently dropped.

   The R flag is the Router flag, and is used by Neighbor Unreachability
   Detection (NUD).  When set, it indicates that the sender is a router.
   An attacker could forge a Neighbor Advertisement message with the
   Router flag cleared to cause the receiving node to remove the



Gont, et al.             Expires April 25, 2014                [Page 13]

Internet-Draft           ND Security Assessment             October 2013


   impersonated Router from the Default router list.

   The S bit is the Solicited flag.  When set it indicates that the
   Neighbor Advertisement is sent in response to a Neighbor Solicitation
   sent from the Destination Address.  The S flag is used as
   reachability confirmation for Network Unreachability Detection (NUD).
   As stated in Section 4.4 of [RFC4861], it must not be set in
   multicast advertisements or in unsolicited unicast advertisements.

   A node that receives a Neighbor Advertisement message that has the
   S-bit set and was sent to a multicast address should silently discard
   the received message.  Additionally, a node that receives an
   unsolicited Neighbor Advertisement message (i.e., there was not a
   pending Neighbor Solicitation for the Target Address) with the S-bit
   set that was sent to a unicast address should silently drop the
   received message.

   The O bit is the Override flag.  When set, it indicates that this
   Neighbor Advertisement should override an existing cache entry and
   update the cached link-layer address.  When the O bit is not set, the
   advertisement will not update a cached link-layer address, but will
   update a Neighbor Cache entry that does not include a link-layer
   address.

   The O bit should be set in all solicited advertisements, except those
   for anycast addresses.  A node that receives an unsolicited Neighbor
   Advertisement message with the O bit set should silently drop the
   received message.  However, we note that it is virtually impossible
   to enforce this requirement for Neighbor Advertisement messages for
   anycast addresses that have the O bit set, as anycast addresses are
   syntactically indistinguishable from normal unicast addresses.

   For solicited Neighbor Advertisements, the Target Address is set to
   the Target Address of the Neighbor Solicitation message that elicited
   the advertisement.  For unsolicited Neighbor Advertisements, the
   Target Address is set to the address whose link-layer address has
   changed.

   The Target Address must pass the following checks:

   1.  It must not be a multicast address (as required in Section 4.4 of
       [RFC4861]

   2.  It must not be the unspecified address (::)

   3.  It must not be the loopback address (::1)

   As of this writing, the following options are allowed in Neighbor



Gont, et al.             Expires April 25, 2014                [Page 14]

Internet-Draft           ND Security Assessment             October 2013


   Advertisement messages:

   o  Target link-layer address

   Other options present in a Neighbor Advertisement should be ignored.

   The target link-layer address specifies the link-layer address of the
   target of the Neighbor Advertisement.  According to Section 4.4 of
   [RFC4861], this option must be included in Neighbor Advertisements
   when they are sent in response to neighbor solicitations sent to
   multicast addresses (provided the link layer has addresses).  A node
   that receives a Neighbor Advertisement message in response to a
   solicitation sent to a multicast address, without a target link-layer
   address should silently drop the received message (provided that the
   corresponding link layer has addresses).

   Section 3.6.3 contains further validation checks that should be
   performed on target link-layer address options.

3.5.  Redirect message

   Routers send Redirect packets to inform a host of a better first-hop
   node on the path to a destination, or to inform a host that the
   destination node is in fact a neighbor (i.e., it is attached to the
   same link).

   The following figure illustrates the syntax of the Redirect message:
























Gont, et al.             Expires April 25, 2014                [Page 15]

Internet-Draft           ND Security Assessment             October 2013


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Code      |          Checksum             |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Reserved                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     +                                                               +
     |                                                               |
     +                       Target Address                          +
     |                                                               |
     +                                                               +
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     +                                                               +
     |                                                               |
     +                     Destination Address                       +
     |                                                               |
     +                                                               +
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |   Options ...
     +-+-+-+-+-+-+-+-+-+-+-+-

                 Figure 5: ICMPv6 Redirect message format

   The Source Address of the IPv6 header is set to the link-local
   address assigned to the interface from which the Redirect message is
   sent.  A node that receives a Redirect message should verify that the
   Source Address of the IPv6 header is a link-local address.  If the
   packet does not pass this check, the Redirect message should be
   silently dropped.  The Source Address of a Redirect message must
   correspond to the IPv6 address of the current first-hop router for
   the specified ICMPv6 Destination Address (i.e., the IPv6 address
   specified in the Destination Address field of the ICMPv6 Redirect
   message).  If the packet does not pass this check, it should be
   silently dropped.

   The Destination Address of the IPv6 header is set to the Source
   Address of the packet that triggered the Redirect.

   The Target Address specifies an IPv6 address that is a better first
   hop to use for the IPv6 address specified in the Destination Address
   field of the ICMPv6 header.  If the Redirect message is meant to
   indicate that a destination is in fact a neighbor (i.e., it is
   attached to the same link), the Target Address is set to the same



Gont, et al.             Expires April 25, 2014                [Page 16]

Internet-Draft           ND Security Assessment             October 2013


   value as the Destination Address field of the ICMPv6 header.

   When the Redirect indicates a first-hop router, the Target Address
   must be a link-local address (that of the aforementioned 'better
   first-hop router').  A node that receives a Redirect message in which
   the Target Address and the Destination Address are different should
   verify that the Target Address is a link-local address.  If the
   Redirect message does not pass this check, it should be silently
   dropped.

   Additionally, the following checks should be performed on the ICMPv6
   Target Address and the ICMPv6 Destination Address:

   1.  They must not contain a multicast address

   2.  They must not contain the unspecified address (::)

   3.  They must not contain the loopback address (::1)

   If a Redirect message does not pass this check, it should be dropped.

   As of this writing, the following options are legitimate for the
   Redirect message:

   o  Target link-layer address

   o  Redirected header

   [RFC4861] specifies that the target-link layer address should be
   included (if known) in Redirect messages, and that it must be
   included for NBMA links that rely on the presence of the Target link-
   layer address option to determine the link-layer address of
   neighbors.

   As explained in Section 8.3 of [RFC4861], if a Redirect message
   contains a Target link-layer address option, the node processing the
   redirect will create or update the Neighbor Cache entry for the
   target.  As a result, an attacker could exploit ICMPv6 Redirect
   messages not only to maliciously update the Destination Cache of the
   victim node, but also (or alternatively) to maliciously update its
   Neighbor Cache.

   The Redirected header option allows the sender of the Redirect
   message to include a portion of the packet that triggered the
   Redirect message.  The Redirected header option is discussed in
   Section 3.6.5.





Gont, et al.             Expires April 25, 2014                [Page 17]

Internet-Draft           ND Security Assessment             October 2013


3.6.  Neighbor Discovery Options

   Neighbor Discovery messages can include a number of options.  Such
   options have the following general syntax:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     |              ...              |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     ~                              ...                              ~
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+


                Figure 6: Neighbor Discovery option format

   The Type field is an 8-bit identifier of the type of option.  As of
   this writing, the following options have been specified:

     +------+---------------------------+----------------------------+
     | Type |          Meaning          |           Summary          |
     +------+---------------------------+----------------------------+
     |   1  | Source link-layer address | Discussed in Section 3.6.2 |
     +------+---------------------------+----------------------------+
     |   2  | Target link-layer address | Discussed in Section 3.6.3 |
     +------+---------------------------+----------------------------+
     |   3  |     Prefix information    | Discussed in Section 3.6.4 |
     +------+---------------------------+----------------------------+
     |   4  |     Redirected header     | Discussed in Section 3.6.5 |
     +------+---------------------------+----------------------------+
     |   5  |            MTU            | Discussed in Section 3.6.6 |
     +------+---------------------------+----------------------------+
     |  24  |     Route Information     | Discussed in Section 3.6.7 |
     +------+---------------------------+----------------------------+
     |  25  |    Recursive DNS Server   | Discussed in Section 3.6.8 |
     +------+---------------------------+----------------------------+
     |  31  |      DNS Search List      | Discussed in Section 3.6.9 |
     +------+---------------------------+----------------------------+

                    Table 1: Neighbor Discovery options

   The Length field specifies the length of the option in units of 8
   octets.  As stated in 4.6 of [RFC4861] a Length of 0 is invalid.
   Nodes must silently discard Neighbor Discovery packets that contain
   an option with a Length of 0.





Gont, et al.             Expires April 25, 2014                [Page 18]

Internet-Draft           ND Security Assessment             October 2013


3.6.1.  General issues with Neighbor Discovery options

   The following subsections discuss security issues that apply to all
   Neighbor Discovery options.

   The proposed checks should be performed in addition to any option-
   specific checks proposed in the next sections.

   Processing requirements

   Processing of Neighbor Discovery options consumes CPU resources at
   the processing node.  While the Hop Limit check of the IPv6 header
   encapsulating a Neighbor Discovery message limits potential attackers
   to those attached to the same link as the target node, there's still
   the potential of an on-link system overwhelming a node by sending it
   packets with a surprisingly large number of Neighbor Discovery
   options.

   To reduce the impact of these packets on the system performance, a
   few counter-measures could be implemented:

   o  Rate-limit the number of Neighbor Discovery packets that are
      processed by the system.

   o  Enforce a limit on the maximum number of options to be accepted in
      any Neighbor Discovery message.

   The first check avoids a large number of Neighbor Discovery packets
   to overwhelm the system in question.  The second check avoids packets
   with multiple Neighbor Discovery options to affect the performance of
   the system.

      Most implementations fail to rate-limit ND packets, and hence have
      been found vulnerable to the aforementioned issue (see e.g.
      [CVE-2011-2391]).

   Option Length

   The Length field specifies the length of the option in units of 8
   octets.  As stated in 4.6 of [RFC4861] a Length of 0 is invalid.
   Nodes must silently discard Neighbor Discovery packets that contain
   an option with a Length of 0.  This check prevents, among other
   things, loops in option processing that may arise from incorrect
   option lengths.

   Additionally, while the Length byte of a Neighbor Discovery option
   allows for an option length of up to 2040 octets (255 * 8 octets),
   there is a limit on legitimate option length imposed by the syntax of



Gont, et al.             Expires April 25, 2014                [Page 19]

Internet-Draft           ND Security Assessment             October 2013


   the IPv6 header.

   For all Neighbor Discovery options, the following check should be
   enforced:

       option-offset + Length * 8 - MIN_IPV6_HEADER <= Payload Length

   Where

   option-offset is the offset of the first byte of the option within
   the IPv6 packet (with the first octet of the IPv6 header having an
   'offset' of 0).  Length is the Length field of the Neighbor Discovery
   option being processed.  MIN_IPV6_HEADER is the size of the fixed
   IPv6 header.  That is, 40 octets.  Payload Length is the header field
   from the IPv6 header encapsulating the Neighbor Discovery message.

   If a Neighbor Discovery option does not pass this check, the
   corresponding Neighbor Discovery message should be silently dropped.

   The aforementioned check is meant to detect forged option-length
   values that might make an option illegitimately exceed the actual
   length of the IPv6 packet encapsulating the Neighbor Discovery
   message.

3.6.2.  Source Link-Layer Address Option

   The Source link-layer address option contains the link-layer address
   of the sender of the packet.  It is used by Neighbor Solicitation,
   Router Solicitation, and Router Advertisement messages.

   The following figure illustrates the syntax of the source link-layer
   address:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     |    Link-Layer Address ...
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

               Figure 7: ND Source link-layer address option

   The Type field is set to 1.  The Length field specifies the length of
   the option (including the Type and Length octets) in units of 8
   octets.  A node that receives an ICMPv6 message with this option
   should verify that the Length field is valid for the underlying link
   layer.  For example, for IEEE 802 addresses the Length field must be
   1 [RFC2464].  If the packet does not pass this check, it should be



Gont, et al.             Expires April 25, 2014                [Page 20]

Internet-Draft           ND Security Assessment             October 2013


   silently dropped.

   The Link-Layer Address field contains the link-layer address.  The
   length, contents, and format of this field varies from one link layer
   to another, and is specified in specific documents that describes how
   IPv6 operates over different link layers.

   A number of validation checks should be performed on the Link-Layer
   Address.  In the case of IEEE 802 addresses, it should not contain a
   broadcast or multicast address.  If the option does not pass this
   check, the Neighbor Discovery message carrying the option should be
   discarded.

   Additionally, nodes should not allow the source link-layer address to
   contain one of the receiving node's link-layer addresses.  If the
   option does not pass this check, the Neighbor Discovery message
   carrying the option should be discarded.

   The source link-layer address option could be exploited for the
   purpose of 'Neighbor Cache poisoning', that is, to cause traffic
   meant for a specific IPv6 address to be illegitimately directed to
   the node whose link-layer address is specified by the Link-Layer
   Address field.

   This is similar to the ARP cache poisoning attacks in IPv4.

   A possible counter-measure for Neighbor Cache poisoning attacks would
   be to override the link-layer address stored in the Neighbor Cache
   only after Neighbor Unreachability Detection (NUD) finds the neighbor
   to be unreachable and the corresponding entry is removed.  This is
   clearly a trade-off between responsiveness and resiliency.

   In some network scenarios it may be possible and desirable to
   configure static Neighbor Cache entries, such that Neighbor Discovery
   need not be performed for the corresponding IPv6 addresses.

   Some implementations have been found to inadvertently override static
   entries when they receive source link-layer address options or target
   link-layer address options in Neighbor Discovery messages
   [Hogg-Vyncke] [Lecigne-Neville-Neil].

   If source link-layer address options were allowed to contain
   broadcast (e.g., the IEEE 802 'ff:ff:ff:ff:ff:ff' address) or
   multicast (e.g., the IEEE 802 '33:33:00:00:00:01' address) addresses,
   traffic directed to the corresponding IPv6 address would be sent to
   the broadcast or multicast address specified in the source link-layer
   option.  This could have multiple implications:




Gont, et al.             Expires April 25, 2014                [Page 21]

Internet-Draft           ND Security Assessment             October 2013


   It would have a negative impact on the performance of the nodes
   attached to the network and on the network itself, as packets sent to
   these addresses would need to be delivered to multiple nodes (and
   processed by them) unnecessarily.

   An attacker could capture network traffic sent to the corresponding
   IPv6 address, as the corresponding packets would be delivered to all
   (in the case of broadcast) or multiple (in the case of multicast)
   nodes.

   Packets could result in forwarding loops at routers, as a router
   forwarding a packet to the corresponding address would receive itself
   a copy of the forwarded packet, thus resulting in a forwarding loop.
   The loop would end only when the Hop Limit is eventually decremented
   to 0.  If multiple routers are present on the same link, the problem
   is further exacerbated.  Section 6.1.10 of this document contains
   further analysis of this vulnerability.

   [Lecigne-Neville-Neil] reports that at least some versions of FreeBSD
   are vulnerable to these issues.

3.6.3.  Target Link-Layer Address Option

   The Target link-layer address option contains the link-layer address
   of the Target of the packet.  It is used by Neighbor Advertisement
   and Redirect messages.

   The following figure illustrates the syntax of the Target link-layer
   address:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     |    Link-Layer Address ...
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

           Figure 8: ND Target link-layer address option format

   The Type field is set to 2.  The Length field specifies the length of
   the option (including the Type and Length octets) in units of 8
   octets.  A node that receives an ICMPv6 message with this option
   should verify that the Length field is valid for the underlying link-
   layer.  For example, for IEEE 802 addresses the Length field must be
   1 [RFC2464].  If the packet does not pass this check, it should be
   silently dropped.

   The Link-Layer Address field contains the link-layer address.  The



Gont, et al.             Expires April 25, 2014                [Page 22]

Internet-Draft           ND Security Assessment             October 2013


   length, contents, and format of this field varies from one link layer
   to another, and is specified in specific documents that describes how
   IPv6 operates over different link layers.

   The target link-layer address has the same security implications as
   the source link-layer address.  Therefore, the same considerations
   apply, and the same validation checks should be performed as for the
   source link-layer address (see Section 3.6.2).

3.6.4.  Prefix Information

   The Prefix Information option is used by routers to provide hosts
   with on-link prefixes and prefixes for Address Auto-configuration.
   It may only appear in Router Advertisement messages and should be
   silently ignored in any other messages [RFC4861].

   The following figure illustrates the syntax of the Prefix Information
   option:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     | Prefix Length |L|A|R|Reserved1|
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                         Valid Lifetime                        |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                       Preferred Lifetime                      |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Reserved2                           |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     +                                                               +
     |                                                               |
     +                            Prefix                             +
     |                                                               |
     +                                                               +
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

               Figure 9: ND Prefix Information option format

   The Type field is set to 3.  The Length field is set to 4 by the
   sender.  A node processing a Prefix Information option should verify
   that the Length field is 4.  If the option does not pass this check,
   the option should be ignored.

   The Prefix Length is an 8-bit unsigned integer that specifies the



Gont, et al.             Expires April 25, 2014                [Page 23]

Internet-Draft           ND Security Assessment             October 2013


   prefix length, that is, the number of leading bits in the Prefix
   field that are valid.

   The following sanity check should be applied on the Prefix Length
   field:

                            Prefix Length >= 32

   If the Prefix Length field does not pass this checks, the Prefix
   Information option should be discarded.

   The L bit is a 1-bit flag that, when set, states that the prefix can
   be used for on-link determination.  The A bit is a 1-bit autonomous
   address-configuration flag that indicates whether this prefix can be
   used for autonomous address configuration.  The R flag is specified
   by [RFC6275], and indicates that the Prefix field contains a complete
   IPv6 address assigned to the sending router.  The Reserved1 field is
   a 6-bit unused field that is set to zero by the sender and must be
   ignored by the receiver.

   The Valid Lifetime field is a 32-bit unsigned integer that specifies
   the amount of time (in seconds) this prefix can be used for on-link
   determination (with a value of 0xffffffff representing 'infinity').
   We recommend hosts to sanitize the Valid Lifetime as follows:

           SanitizedVL = max(Valid Lifetime, MIN_VALID_LIFETIME)

   Where SanitizedVL is the sanitized 'Valid Lifetime', and
   MIN_VALID_LIFETIME is set to 1800 (seconds).

   The value of 1800 seconds for MIN_VALID_LIFETIME has been selected to
   coincide with the lower limit enforced on the Router Lifetime
   (MIN_ROUTER_LIFETIME).

   The Preferred Lifetime is a 32-bit unsigned integer that specifies
   the length of time (in seconds) that addresses generated from this
   prefix via stateless address auto-configuration (SLAAC) should remain
   'preferred' (with a value of 0xffffffff representing 'infinity').

   As noted in [RFC4861] the Preferred Lifetime must be smaller than or
   equal to the Valid Lifetime to avoid preferring addresses that are no
   longer valid.  Therefore, a node processing a Prefix Information
   option should perform the following check:

                    Preferred Lifetime <= Valid Lifetime

   If the option does not pass this check, it should be silently
   ignored.



Gont, et al.             Expires April 25, 2014                [Page 24]

Internet-Draft           ND Security Assessment             October 2013


   The Reserved2 is a 32-bit unused field that is set to zero by the
   sender and must be ignored by the receiver.

   The Prefix field contains an IPv6 address or a prefix of an IPv6
   address.

   The Prefix Length contains the number of leading bits in the prefix
   that are to be considered valid.  The remaining bits in the Prefix
   field are set to zero by the sender and must be ignored by the
   receiver.

   As stated in Section 4.6.2 of [RFC4861], routers should not send a
   Prefix Information option for the link-local prefix.  Therefore, a
   node should verify that the Prefix does not contain the link-local
   prefix.  If the option does not pass this check, it should be
   silently dropped.

   Additionally, a node should verify that the Prefix does not contain a
   multicast IPv6 prefix.  If the option does not pass this check, it
   should be silently dropped.

   An attacker could exploit the Prefix information option to perform a
   Denial-of-Service attack, by sending a large number of Router
   Advertisements with the Prefix Information options that have the A
   bit set, therefore advising the receiving systems to configure an
   IPv6 address with each of these prefixes.  If an implementation does
   not enforce a limit on the number of addresses they configure in
   response to Router Advertisements, the aforementioned attack might
   result in buffer overflows or kernel memory exhaustion.

      [CVE-2010-4669] is one vulnerability report about the
      aforementioned issue.

   We recommend hosts to default to a maximum number of configured
   addresses (for each interface) of 16.

   This limits is already being enforced by a number of implementations,
   such as OpenBSD 4.2.

   On the other hand, Windows XP SP2 and FreeBSD 9.0 fail to enforce
   limits on the maximum number of configured addresses, and therefore
   are vulnerable to a Denial of Service attack.

   Even if hosts do enforce a limit on the number of IPv6 addresses
   configured, an attacker might try to cause victim hosts to ignore
   legitimate prefixes previously advertised for address configuration
   by legitimate routers.  Hereby we recommend hosts to not discard
   previously configured addresses if new prefixes for address auto-



Gont, et al.             Expires April 25, 2014                [Page 25]

Internet-Draft           ND Security Assessment             October 2013


   configuration are advertised and the limit for the maximum number of
   configured addresses (per interface) has been reached.  When such
   limit is hit, the newly advertised prefixes for address auto-
   configuration should be ignored.

   Section 3.6.4 describes how an attacker could exploit the Prefix
   Information option for the purpose of traffic hijacking.

3.6.5.  Redirected Header Option

   The Redirected Header option is used in Redirect messages to convey
   all or part of the packet that is being redirected.  It must be
   silently ignored for all other messages.

   The following figure illustrates the syntax of the Redirected Header
   option:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     |            Reserved           |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Reserved                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     ~                       IP header + data                        ~
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

               Figure 10: ND Redirected Header option format

   The Type field is 4.  The Length field specifies the option size
   (including the Type and Length fields) in units of 8 octets.
   Assuming the Redirected Header option will contain at least the
   mandatory fields of the option (8 bytes), the fixed IPv6 header (40
   bytes), and the first 8 bytes of the transport protocol header, the
   following validation check should be performed:

                                Length >= 7

   If the option does not pass this check, the corresponding Redirect
   Message should be silently ignored.

   As the option is meant to contain as much of the Redirected packet
   without exceeding the minimum IPv6 MTU, and the minimum IPv6 MTU is
   1280 octets, this is a sensible requirement to enforce.




Gont, et al.             Expires April 25, 2014                [Page 26]

Internet-Draft           ND Security Assessment             October 2013


3.6.6.  MTU Option

   The MTU option is used in Router Advertisement messages to ensure
   that all nodes on a link use the same MTU value in those scenarios in
   which heterogeneous technologies are bridged together.  This option
   must be silently ignored for other Neighbor Discovery messages.

   The following figure illustrates the syntax of the MTU option:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     |           Reserved            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                              MTU                              |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                      Figure 11: ND MTU option format

   The Type field identifies the kind of option and is set to 5.  This
   option has a fixed Length that is 1.  Therefore, the following sanity
   check should be performed:

                                Length == 1

   If the option does not pass this check it, should be ignored.

   The Reserved field is a 16-bit unused field that is set to 0 by the
   sender and should be ignored by the receiver.

   The MTU field is a 32-bit unsigned integer that specifies the MTU
   value that should be used for this link.  [RFC2460] specifies that
   the minimum IPv6 MTU is 1280 octets.  Therefore, a node processing a
   MTU option should perform the following check:

                          MTU >= MINIMUM_IPV6_MTU

   where MINIMUM_IPV6_MTU is a variable that should be set to 1280.

   If the option does not pass this check, it should be silently
   ignored.

   It has been reported that some link layers do not support a minimum
   MTU of 1280 bytes.  Therefore, implementations should provide the
   means to change the default value of the MINIMUM_IPV6_MTU variable.

   Additionally, the advertised MTU should not exceed the maximum MTU



Gont, et al.             Expires April 25, 2014                [Page 27]

Internet-Draft           ND Security Assessment             October 2013


   specified in the link-type-specific document (e.g., [RFC2464] for
   Ethernet networks).  Therefore, a node processing a MTU option should
   perform the following check:

                            MTU <= MAX_LINK_MTU

   where MAX_LINK_MTU is a variable that should be set according to the
   maximum link MTU specified in the link-type-specific document (e.g.,
   [RFC2464] for Ethernet).

   If the option does not pass this check, it should be silently
   ignored.

   The MTU option could be potentially exploited by an attacker to
   perform a Denial-of-Service (DoS) or a performance-degrading attack
   against the systems in a local network.  In order to perform this
   attack, an attacker would forge a Router Advertisement that includes
   an MTU option with a very small (possibly zero) value.  The impact of
   this attack would be the same as the 'Blind performance-degrading
   attack' described in Section 15.7 of [CPNI-TCP].

3.6.7.  Route Information Option

   The Route Information option is used to convey more-specific routes
   in Router Advertisement messages.  The following figure illustrates
   the syntax of the Router Information option:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |    Length     | Prefix Length |Resvd|Prf|Resvd|
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                        Route Lifetime                         |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                   Prefix (Variable Length)                    |
     .                                                               .
     .                                                               .
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

               Figure 12: ND Route Information option format

   The Type field identifies the type of option and is set to 24.  The
   Length field contains the length of the option (including the Type
   and Length fields) in units of 8 octets.  The following sanity checks
   should be performed on these Length field:

                        (Length >=1) && (Length <=3)



Gont, et al.             Expires April 25, 2014                [Page 28]

Internet-Draft           ND Security Assessment             October 2013


   If the option does not pass this check, it should be ignored.

   An option Length of 1 octet allows the specification of prefixes with
   a length of 0 (i.e., /0), while an option Length of 3 allows the
   specification of prefixes of up to 128 bits (i.e., /128).

   The Prefix Length field indicates the number of leading bits in the
   Prefix field that are valid.  The Length field and the Prefix Length
   field are closely related, as the Length field constrains the
   possible values of the Prefix Length field.

   The following sanity check should be enforced on the Prefix Length
   field:

                     Prefix Length <= (Length - 1) * 64

   If the option does not pass this check, it should be ignored.

   Both of the Rsvd fields are set to zero by the sender of the message,
   and should be ignored by the receiver.  The Prf field specifies the
   'Preference' of this route.  As specified by RFC 4191, if the Prf
   field contains the value of '10' ('Reserved'), the option should be
   ignored.

   The Route Lifetime field specifies the length of time, in seconds,
   that the prefix is valid for route determination.  While not required
   by RFC 4191, we recommend hosts to perform the following sanity check
   on the Route Lifetime field:

        SanitizedRL = min( max(Route Lifetime, MIN_ROUTE_LIFETIME),
                           MAX_ROUTE_LIFETIME)

   Where MIN_ROUTE_LIFETIME is set to 1800 seconds and
   MAX_ROUTE_LIFETIME is set to 9000 seconds.

   These values have been selected according to the upper and lower
   limits described in Section 3.2 ('Router Advertisement') of this
   document for the Router Lifetime field of Router Advertisements.

   The Prefix field contains the actual IPv6 prefix that, together with
   the Prefix Length field, identifies the route.  A number of sanity
   checks should be enforced on the Prefix field.  For example, the
   Prefix should neither contain a link-local unicast prefix (e.g.,
   fe80::/10, fe80::/64, etc.) nor a link-local multicast prefix (e.g.,
   ff02::0/64).

   The Route information option has a number of security implications.
   Firstly, an attacker could forge Router Advertisements with a higher



Gont, et al.             Expires April 25, 2014                [Page 29]

Internet-Draft           ND Security Assessment             October 2013


   'preference' value (Prf), thus overriding the existing default
   routers at the attacked system.  Secondly, an attacker could exploit
   this option to implant more specific routes to a victim prefix at the
   attacked system, thus overriding the existing routes for that victim
   prefix.  Thirdly, an attacker could cause an existing route to a
   victim prefix to be removed from the routing table of the attacked
   host, by forging a Route Information option that contains a Route
   Lifetime of 0 (or some other small value).  Fourthly, if an
   implementation does not enforce limits on the size of the Destination
   Cache, an attacker could possibly exhaust the kernel memory or CPU
   cycles of that system by forging a large number of Route Information
   options (possibly in many different Router Advertisements), such that
   the attacked system consumes its kernel memory and/or its CPU time to
   install those routes (see e.g.  [CVE-2012-notyet]).

   A general mitigation for the security implications of Router
   Advertisements that can be applied when the protocols are deployed is
   to restrict which ports of a managed switch can send Router
   Advertisement messages.  That is, Router Advertisements received on
   all other ports of the switch would be discarded.  This mechanism is
   commonly-known as Router Advertisement Guard (RA-Guard) [RFC6104]
   [RFC6105] [I-D.ietf-v6ops-ra-guard-implementation].

   We recommend hosts to not simply discard a default router entry when
   a Router Preference with a higher Prf value is received.  In
   particular, default routers that are known to be working should not
   be discarded when such Router Advertisements are received.

      This means that the higher-priority router would override the
      existing default router, but the latter would still be kept in the
      "default routers list", such that if the newly-learned router is
      found to be non-working, the existing (lower-priority) router
      could still be employed).

   We recommend hosts to enforce a lower limit Prefix Length in the
   Route Information options.  This would prevent an attacker from
   overriding the default routers by including, e.g., one Route
   Information option for the prefix ::/1 and one Route Information
   option for the prefix 8000::/1.  We recommend hosts to enforce a
   minimum Prefix Length of 32.  Hosts may also enforce an upper limit
   on the Prefix Length, such that an attacker cannot easily redirect
   traffic to specific site.  A possible upper limit for the Prefix
   Length would be 64.  As discussed earlier in this Section, the Route
   Lifetime value should be sanitized, and a route entry should not
   simply be discarded when a Route Information option with a Route
   Lifetime of 0 is received.

   Finally, hosts should enforce a limit on the maximum number of



Gont, et al.             Expires April 25, 2014                [Page 30]

Internet-Draft           ND Security Assessment             October 2013


   entries in the Destination Cache.

3.6.8.  Recursive DNS Server Option

   The Recursive DNS Server (RDNSS) option provides a mechanism for
   routers to advertise the IPv6 addresses of one or more Recursive DNS
   Servers.  This option is specified in [RFC6106].  The following
   figure illustrates the syntax of the RDNSS options:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Length    |           Reserved            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Lifetime                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     :            Addresses of IPv6 Recursive DNS Servers            :
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

             Figure 13: ND Recursive DNS Server option format

   The Type field identifies must be 25.  The Length field specifies the
   length of the option (including the Type and Length fields) in units
   of 8 octets.  The following sanity checks should be performed on the
   Length field:

                                Length >= 3

   If the option does not pass this check, it should be ignored.

   This sanity check requires a RDNSS option to contain the IPv6 address
   of at least one Recursive DNS Server.

   Additionally, the following sanity check should be performed:

                            (Length -1) % 2 == 0

   If the option does not pass this check, it should be silently
   ignored.

   As an IPv6 address consists of 16 bytes, each IPv6 address that is
   included in the option should increment the minimum option length by
   2.

   The Reserved field is set to zero by the sender, and must be ignored



Gont, et al.             Expires April 25, 2014                [Page 31]

Internet-Draft           ND Security Assessment             October 2013


   by the receiver.  The Lifetime field specifies the maximum time in
   seconds that a node may use the IPv6 addresses included in the option
   for name resolution, with a value of 0 indicating that they can no
   longer be used.  As specified in [RFC6106], the following sanity
   checks should be performed on the Lifetime field:

                              Lifetime >= 1800


                              Lifetime <= 3600

      [RFC6106] specifies these sanity checks as MaxRtrAdvInterval <=
      Lifetime <= 2*MaxRtrAdvInterval.

   If the Lifetime field does not pass this check, the option should be
   ignored.

   Failure to enforce a lower limit on the Lifetime value could allow an
   attacker to 'disable' a Recursive DNS Server at a target system, by
   forging a Router Advertisement with a RDNSS option that includes the
   IPv6 address of such DNS Server, and a Lifetime of 0 (or some other
   small value).  This would cause the receiving system to remove such
   RDNSS address from the list of Recursive DNS Servers.  However, it
   should be noted that this represents a trade-off of responsiveness
   vs. resiliency.

   Sanity checks should be performed on the IPv6 addresses that are
   included in the RDNSS option.  For example, nodes should check that
   the IPv6 addresses included in the RDNSS option are not multicast
   addresses.  If any of the addresses in the RDNSS option does not pass
   this check, it should be silently dropped.

   Nodes should enforce a limit on the number of IPv6 addresses they
   include in the local list of Recursive DNS Servers.  An arbitrary
   limit could be to allow a maximum of 16 IPv6 addresses in the list of
   Recursive DNS Servers.

   The value 16 is somewhat arbitrary.  It has been chosen to be the
   same as the limit on the maximum number of default routes that many
   systems (such as OpenBSD 4.2) already enforce.

   Failure to enforce limits on the maximum number of Recursive DNS
   Servers could probably allow an attacker to exhaust the system memory
   by crafting multiple Router Advertisements that advertise a large
   number of IPv6 addresses in RDNSS options.

   It should also be clear that if an attacker is able to advertise a
   malicious Recursive DNS Server to victim nodes, he could perform a



Gont, et al.             Expires April 25, 2014                [Page 32]

Internet-Draft           ND Security Assessment             October 2013


   variety of attacks against the victim nodes (DoS, Man-in-the-Middle.
   Etc.).

3.6.9.  DNS Search List

   The Recursive DNS Server (RDNSS) option provides a mechanism for
   routers to advertise the IPv6 addresses of one or more Recursive DNS
   Servers.  This option is specified in [RFC6106].  The following
   figure illustrates the syntax of the RDNSS options:


      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
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |     Type      |     Length    |           Reserved            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                           Lifetime                            |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |                                                               |
     :                Domain Names of DNS Search List                :
     |                                                               |
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                               Figure 14: V

   XXX (PLACEHOLDER): Need to complete the security assessment of this
   option.
























Gont, et al.             Expires April 25, 2014                [Page 33]

Internet-Draft           ND Security Assessment             October 2013


4.  Router and Prefix Discovery

4.1.  Router Specification

   Section 6.2 of [RFC4861] contains the Router specification for Router
   and Prefix Discovery.

   Section 6.2.6 ('Processing Router Solicitations') of [RFC4861] states
   that if a router receives a Router Solicitation message, and 'the
   router already has a Neighbor Cache entry for the solicitation's
   sender, the solicitation contains a Source Link-Layer Address option,
   and the received link-layer address differs from that already in the
   cache, then the link-layer address SHOULD be updated in the
   appropriate Neighbor Cache entry'.  As a result, an attacker might
   forge a Router solicitation message with a forged source link-layer
   address, thus causing all traffic meant from the attacked router to
   the (forged) Source Address of the Router Advertisement to be sent to
   the link-layer address advertised in the forged source link-layer
   address option.

   Section 6.2.6 of [RFC4861] further states that 'Whether or not a
   Source Link-Layer Address option is provided, if a Neighbor Cache
   entry for the solicitation's sender exists (or is created) the
   entry's IsRouter flag MUST be set to FALSE'.  As a result, in a
   network scenario in which there are two routers ('A' and 'B') on the
   same link, and an attacker is directly attached to that link, an
   attacker could send a Router Solicitation to one of the routers
   (Router A) forging the Source Address to be that of the other router
   (Router B).  As a result, the target router (Router A) would set the
   IsRouter flag of the Neighbor Cache entry corresponding to the IPv6
   address of Router B (the forged Source Address of the Router
   Solicitation message) to FALSE, and as a result, Router B would be
   eliminated from the Default router list of Router A.

   One interesting aspect about this attack is that while some devices
   might be filtering e.g.  Router Advertisements, they might not be
   filtering Router Solicitation messages, and thus this attack might
   still be effective.

4.2.  Host Specification

   Section 6.3.4 of [RFC4861] states that when a Router Advertisement is
   received that communicates information for a specific parameter
   (e.g., link MTU) that differs from information received in previous
   Router Advertisements, the most recently received information is
   considered authoritative.

   While this requirement guarantees that the relevant information can



Gont, et al.             Expires April 25, 2014                [Page 34]

Internet-Draft           ND Security Assessment             October 2013


   be updated in a timely fashion, it also guarantees that an attacker
   connected to the local link always has the chance to maliciously
   override the values of parameters previously learned from Router
   Advertisements.

   Section 6.3.4 of [RFC4861] states that 'to limit the storage needed
   for the Default Router List, a host MAY choose not to store all of
   the router addresses discovered via advertisements'.  Here we
   strongly advise hosts to enforce a limit on the maximum number of
   entries in the Default Router List.  A possible (somewhat arbitrary)
   limit for the maximum number of entries in the Default Router list
   would be 16.

   Section 6.3.4 of [RFC4861] states that 'If the received Cur Hop Limit
   value is non-zero, the host SHOULD set its CurHopLimit variable to
   the received value'.  Here we strongly advise that the received Cur
   Hop Limit is sanitized as described in Section 3.2 of this document.

   Section 6.3.4 of [RFC4861] states that 'The RetransTimer variable
   SHOULD be copied from the Retrans Timer field, if the received value
   is non-zero'.  Here we strongly advise that the received Retrans
   Timer is sanitized as described in Section 3.2 of this document.

   Honouring very small Retrans Timer values could lead the system to
   flood the network with Neighbor Advertisements.

   With respect to the processing of received MTU options, Section 6.3.4
   of [RFC4861] correctly states that the received option should be
   honoured as long as the received value is within the expected limits.
   Section 3.6.6 of this document discusses a number of checks that
   should be performed on received MTU options.

   Section 6.3.4 of [RFC4861] states that 'The only way to cancel a
   previous on-link indication is to advertise that prefix with the
   L-bit set and the Lifetime set to zero'.  This means that if an
   attacker forges a Router Advertisement that advertises a 'victim'
   prefix as being on-link, such prefix will usually be considered 'on-
   link' for the advertised Lifetime period of time ('forever' if
   Lifetime was set to 0xffffffff).












Gont, et al.             Expires April 25, 2014                [Page 35]

Internet-Draft           ND Security Assessment             October 2013


5.  Address Resolution

   In the case of broadcast link-layer technologies, in order for a
   system to transfer an IPv6 datagram, it must first map an IPv6
   address to the corresponding link-layer address via Neighbor
   Solicitation/Advertisement messages.

   While this operation is being performed, the packets that require
   such a mapping would need to be kept in memory.  This may happen both
   in the case of hosts and in the case of routers.

   The possible implementation approach of keeping those datagrams in
   memory while the mapping operation is being performed is mentioned in
   Section 5.2 of [RFC4861].

   This situation might be exploited by an attacker to perform a Denial-
   of-Service (DoS) attack against an IPv6 router, by sending a large
   number of packets to a non-existent node that would supposedly be a
   neighbor to the attacked router.  While trying to map the
   corresponding IPv6 address into a link-layer address, the attacked
   router would keep in memory all the packets that would depend on that
   address resolution operation in order to be forwarded to the intended
   destination.  At the point in which the mapping function times out,
   the node would typically drop all the packets that were queued
   waiting for that address resolution operation to complete.

   Depending on the timeout value for the mapping function this
   situation might result in the attacked router running out of memory,
   with the consequence that incoming legitimate packets would have to
   be dropped, or that legitimate packets already stored in the router's
   memory buffers might need to be dropped.  Both of these situations
   would lead either to a complete Denial of Service or to a degradation
   of the network service.

   A number of countermeasures are warranted for this vulnerability:

   Firstly, nodes should enforce a limit on the maximum number of
   packets that are queued for the same destination address while the
   corresponding address resolution operation is being performed.
   Additionally, nodes should enforce a limit on the aggregate number of
   packets that are queued waiting for address resolution operations to
   complete.

   At the point the mapping function times out, all the packets destined
   to the address that timed out should be discarded.  In addition, a
   'negative cache entry' might be kept in the module performing the
   matching function, so that for some amount of time the mapping
   function would return an error when the IPv6 module requests



Gont, et al.             Expires April 25, 2014                [Page 36]

Internet-Draft           ND Security Assessment             October 2013


   resolution of some IPv6 address for which the mapping has recently
   failed (i.e., timed out).

   A proactive mitigation for this vulnerability would be that when a
   packet is received that requires an address resolution operation
   before the packet can be forwarded, the packet is dropped and the
   router is then engaged in the address resolution operation.

   This is a common implementation strategy for IPv4 routers.  In IPv4,
   it is common that when a packet is received that requires an ARP
   request to be performed (before the packet can be forwarded), the
   packet is dropped and the router is then engaged in the ARP
   procedure.

   While similar issues exist in IPv4 networks, this problem is
   exacerbated in IPv6, as IPv6 subnets are typically much larger than
   their IPv4 counterparts.  Therefore, an attacker could send a large
   number of packets, each destined to different IPv6 addresses
   corresponding to non-existent 'neighbor nodes' of the attacked
   router.  In the event that the router implementation drops packets
   only when the address resolution operation times out, the DoS
   condition might persist, whereas it would have probably disappeared
   if all the malicious packets had been destined to the same IPv6
   address.

   That is, if all the attack packets had been destined to the same IPv6
   address, the timeout of the address resolution operation for that
   IPv6 address could have resulted in all the attack packets to be
   dropped.

   An attacker could also potentially perform a Denial-of-Service attack
   against a router by exhausting the number of entries in the Neighbor
   cache and/or the Destination cache.  In order to perform this attack,
   an attacker would send a large number of packets, each destined to
   different IPv6 addresses corresponding to non-existent 'neighbor
   nodes' of the attacked router.  Each of these attack packets would
   trigger an address-resolution operation at the attacked router.  If
   the target router does not enforce any limits on the maximum number
   of entries in the Neighbor cache, this attack could result in a
   buffer overflow at the attacked router.  On the other hand, if the
   router does enforce limits on the maximum number of entries in the
   neighbor cache, the packets sent by the attacker could result in the
   attacked router hitting the aforementioned limit, probably preventing
   legitimate entries to be added to the Neighbor cache, resulting in a
   Denial of Service to the corresponding nodes.

   This situation has been experienced in production networks probably
   as a result of reconnaissance activity by attackers.  That is, this



Gont, et al.             Expires April 25, 2014                [Page 37]

Internet-Draft           ND Security Assessment             October 2013


   situation could not only indicate a deliberate Denial-of-Service
   attack against a router, but could also be side-effect of network
   reconnaissance (i.e., 'scanning') activities.

   A number of mitigations are warranted for this vulnerability:

   o  Nodes should enforce a limit on the number of entries in the
      Neighbor cache.

   o  Nodes should implement an algorithm to reclaim Neighbor Cache
      entries when the limit on the number of entries is reached.

   o  Nodes should enforce a limit on the number of entries in the
      Neighbor Cache that have a reachability state of 'INCOMPLETE'.
      This limit should be much stricter than the general limit on the
      number of entries in the Neighbor Cache.

   o  Nodes should enforce a limit on the number of entries in the
      Destination cache.

   o  Nodes should implement an algorithm to reclaim Destination Cache
      entries when the limit on the maximum number of entries is
      reached.

   Section 5.3 of [RFC4861] states that for the purpose of eliminating
   unused entries (i.e., garbage-collection) in the Neighbor cache, any
   Least Recently Used (LRU)-based policy that only reclaims entries
   that have not been used in some time should be adequate.  Such LRU-
   based policy should also be enforced to reclaim entries in the
   Neighbor cache or the Destination Cache when the limit on the maximum
   number of entries is hit for the Neighbor cache or the Destination
   cache, respectively.

5.1.  Interface initialization

   As explained in Section 7.2.1 of [RFC4861], when a multicast-capable
   interface is enabled, the node must join the all-nodes multicast
   group on that interface, and the solicited-node multicast address
   corresponding to each of the addresses assigned to an interface.  As
   discussed in the same section, nodes join and leave the solicited-
   node groups as the assigned addresses change over time.

   As the solicited-node multicast address for a number of assigned
   addresses might be the same, nodes should ensure that a solicited-
   node multicast group is not left until all the addresses
   corresponding to that solicited-node group have been removed.

   An implementation that incorrectly leaves a solicited-node multicast



Gont, et al.             Expires April 25, 2014                [Page 38]

Internet-Draft           ND Security Assessment             October 2013


   group while there are addresses corresponding to that multicast group
   still in use might be subject of a Denial-of-Service attack (from a
   malicious node attached to the same link as the victim).

   In order to perform such an attack, an attacker would first send an
   unsolicited Router Advertisement, announcing a prefix such that the
   victim node configures an address whose solicited-node multicast
   group is the same as some legitimately-configured address.  The
   advertised prefix would have a Lifetime value that would cause the
   address to be removed in the short term.  If the Neighbor Discovery
   implementation of the victim system does not ensure that a solicited-
   node multicast group is left only when the last address corresponding
   to that solicited-node multicast group is removed, the victim might
   incorrectly leave the aforementioned solicited-node multicast group.
   As a result, the victim system would be unable to respond to Neighbor
   Solicitation messages for the target address, and therefore the
   aforementioned address would become unreachable.

5.2.  Receipt of Neighbor Solicitation messages

   As stated in Section 7.2.3, if a Neighbor Solicitation is received
   and an entry already exists in the Neighbor Cache for the IPv6 Source
   Address of the solicitation with a cached link-layer address that is
   different from the one in the received Source Link-Layer option, the
   cached address should be replaced by the received address (and the
   entry's reachability state must be set to STALE).

   If an entry does not exist for the corresponding Target Address, a
   new entry is added to the Neighbor Cache, and its reachability state
   is set to STALE.

   While this allows for improved responsiveness in the event the link-
   layer address corresponding to an IPv6 address changes, it also means
   that an attacker could easily override the mapping from IPv6 address
   to link-layer address.

   An attacker attached to the same link as the victim system could
   craft a Neighbor Solicitation message with a forged IPv6 Source
   Address, and send it to the victim system, thus illegitimately
   causing the victim to update its Neighbor Cache.  The attacker could
   then send a Neighbor Advertisement with the Solicited flag set, thus
   causing the reachability state of the neighbor cache entry to be set
   to REACHABLE.

   As a result, the attacker could cause traffic meant from the victim
   to the forged IPv6 address to be directed to any local system (i.e.,
   attached to the same network link).




Gont, et al.             Expires April 25, 2014                [Page 39]

Internet-Draft           ND Security Assessment             October 2013


6.  Vulnerability analysis

   This section summarizes the security implications that arise from the
   Neighbor Discovery mechanisms in IPv6.  The following vulnerabilities
   have been identified:

   o  Denial of Service: communication is prevented between legitimate
      nodes

   o  Performance-degrading: the performance of communication between
      legitimate nodes is reduced

   o  Traffic hijacking: traffic is illegitimately redirected to a node
      operated by an attacker

   [RFC3756] provides a good security assessment of the Neighbor
   Discovery mechanisms.  The following subs-sections summarize the
   results in [RFC3756], and also identify new vulnerabilities and
   specific attack vectors not present in that document.

   Some of the vulnerabilities discussed in the following sub-sections
   involve an attacker sending Router Advertisement messages.  [RFC6104]
   analyzes the problem of Rogue IPv6 Router Advertisements, and discuss
   a number of possible solutions.  [RFC6105] discusses a specific
   solution to this problem based on layer-2 filtering, known as 'RA-
   Guard'.  However, as discussed in
   [I-D.ietf-v6ops-ra-guard-implementation], some popular RA-Guard
   implementations can be easily circumvented by leveraging IPv6
   extension headers.

   [SI6-Toolkit] is a complete complete IPv6 toolkit that can be
   employed to exploit all the vulnerabilities discussed in the
   following subsections.  [THC-IPv6] includes 'proof of concept' tools
   of some of the vulnerabilities discussed in the following
   subsections. [vanHauser2006] is a presentation about this tool set.

   [Beck2007b] analyzes the use of Neighbor Discovery for Operating
   System detection.  However, the results seem to indicate that
   Neighbor Discovery is not an attractive means for Operating System
   detection when compared to other techniques such as those described
   in [CPNI-TCP].  Therefore, the 'information leakage' resulting from
   Neighbor Discovery, while possible, is not included in the threat
   analysis present in the following subsections.

6.1.  Denial of Service






Gont, et al.             Expires April 25, 2014                [Page 40]

Internet-Draft           ND Security Assessment             October 2013


6.1.1.  Neighbor Cache poisoning

   Router Solicitation, Router Advertisement, Neighbor Solicitation and
   Neighbor Advertisement messages can be exploited to maliciously
   poison the Neighbor Cache of a victim node such that an IPv6 address
   maps into a non-existent link-layer address.  As a result, traffic
   meant to the victim address would be 'black-holed' as a result of
   sending it to a non-existent link-layer address.

   In the case of Router Solicitation, Router Advertisement, and
   Neighbor Solicitation messages, a source link-layer address would be
   employed.  In the case of Neighbor Advertisement messages, a target
   link-layer address would be used instead.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link to which the
   target node is attached, or control a node attached to that network
   link (e.g., compromise such a node).  However, it could be possible
   that as a result of tunnelling mechanisms, an attacker could perform
   these attacks remotely.

   This attack could be mitigated by not overwriting the link-layer
   address of an existing Neighbor Cache entry when a source link-layer
   address option or a target link-layer address option is received.
   The mapping of IPv6 address to link-layer address would be updated
   only if Neighbor Unreachability Detection (NUD) first removes the
   corresponding Neighbor Cache entry, and then a Neighbor Discovery
   message updates the mapping.

   Furthermore, some monitoring tools exist that detect some possible
   exploitation of Neighbor Discovery and Neighbor Advertisement
   messages.  NDPMon [NDPMon] is an example of such a monitoring tool
   (which is similar to the IPv4 arpwatch tool [arpwatch]).  [Beck2007]
   is a paper about the aforementioned tool.

6.1.2.  Tampering with Duplicate Address Detection (DAD)

   The Duplicate Address Detection (DAD) mechanism is used to ensure
   that a node does configure for itself an address that is already in
   use.

   An attacker could simply listen to Neighbor Solicitations sent as
   part of the DAD mechanism, and respond with crafted Neighbor
   Advertisements, thus causing the victim node to consider the address
   to be already in use, and thus preventing it from configuring the
   address for future use.

   Additionally, an attacker could respond to Neighbor Solicitations



Gont, et al.             Expires April 25, 2014                [Page 41]

Internet-Draft           ND Security Assessment             October 2013


   sent as part of the DAD mechanism with a Neighbor Solicitation for
   the same IPv6 address.  The legitimate node performing DAD would
   consider this a collision and would cease to solicit that address
   (and possibly select and perform DAD for some alternative address).

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   Layer-2 switches could filter Neighbor Advertisement messages based
   on previous knowledge of the link-layer addresses recently in use at
   each port.

6.1.3.  Tampering with Neighbor Unreachability Detection (NUD)

   The Neighbor Unreachability Detection (NUD) mechanism is used to
   detect unreachable neighbors and cause the corresponding entries in
   the Neighbor Cache to be eliminated.  When an unreachable neighbor is
   detected and the corresponding Neighbor Cache entry is removed, a
   node has the chance to e.g., perform next-hop determination.

   In order for a neighbor to be considered reachable, NUD uses
   reachability indications from upper-layer protocols.  In the absence
   of reachability indications from an upper layer (e.g., from TCP's
   Acknowledgements), NUD employs solicited unicast Neighbor
   Solicitations to confirm the reachability of a Neighbor.

   An attacker could snoop traffic and respond to the solicited Neighbor
   Solicitation messages being used for the purpose of NUD, thus
   preventing victim nodes from detecting that the impersonated node is
   unreachable.  As a result, those 'victim' nodes would continue
   sending packets to the unreachable node, instead of e.g., performing
   first-hop determination to find an alternative working router.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   Nodes could require the link-layer source address of solicited
   Neighbor Advertisements being employed for NUD to be the same as that
   stored in the Neighbor Cache entry for which NUD is being performed.
   With this requirement in place, layer-2 switches could filter
   Neighbor Advertisement messages according to their source link-layer
   address, based on previous knowledge of the link-layer addresses
   recently in use at each port.




Gont, et al.             Expires April 25, 2014                [Page 42]

Internet-Draft           ND Security Assessment             October 2013


   It should be noted that this recommendation should not be enforced in
   more complex networks in which VRRP [RFC5798] or custom redundancy
   protocols are employed.

   This would mitigate the tampering with NUD that employs Neighbor
   Advertisement messages.  However, it should be noted that an attacker
   might still tamper with NUD by forging upper-layer packets such as
   TCP Acknowledgements.

6.1.4.  Rogue Router

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising a
   non-existent system as a default router.

   As a result, hosts honouring the aforementioned Router Advertisements
   would use the advertised rogue router as a default router, and as a
   result their packets would be black-holed.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).  As described in [RFC3756], this
   vulnerability could be mitigated by preferring existing routers over
   new ones.

   Additionally, layer-2 switches could possibly allow Router
   Advertisements messages to be sent only from specific ports.

   [RFC6104] analyzes the problem of Rogue IPv6 Router Advertisements,
   and discusses a number of possible solutions.  [RFC6105] discusses a
   specific solution to this problem based on layer-2 filtering, known
   as 'RA-Guard'.  However, as discussed in
   [I-D.ietf-v6ops-ra-guard-implementation], some popular RA-Guard
   implementations can be easily circumvented by leveraging IPv6
   extension headers.  [CVE-2011-2395] is a vulnerability advisory about
   this issue.

      [SI6-Toolkit] is a complete complete IPv6 toolkit that can be
      employed to circumvent the aforementioned RA-Guard
      implementations.

6.1.5.  Parameter spoofing

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising a
   legitimate default router, but malicious network parameters.




Gont, et al.             Expires April 25, 2014                [Page 43]

Internet-Draft           ND Security Assessment             October 2013


   For example, an attacker could advertise a very small Cur Hop Limit
   value, thus causing packets to be discarded before reaching their
   intended destination.

   An attacker could also advertise an incorrect link MTU (either very
   small or very large) possibly preventing packets from being sent on
   the corresponding link and/or causing pathological behaviour at the
   victim nodes.

   Finally, an attacker could either send unsolicited Router
   Advertisements and/or illegitimately respond to Router Solicitations,
   sending Router Advertisements with the M and/or the O bits set, thus
   possibly causing the victim nodes to engage in managed address
   configuration when such service is not present (e.g., there is no
   DHCP server) or when the attacker operates a malicious DHCP server.

   In the former scenario, address configuration would fail, as a result
   of the non-existing DHCP server.  In the latter scenario, an attacker
   could operate a malicious DHCP server to override IPv6's stateless
   configuration.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   As with other attacks based on Router Advertisement messages, they
   could be mitigated if Layer-2 switches allow Router Advertisements
   messages to be sent only from specific ports.

6.1.6.  Bogus on-link prefixes

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising
   bogus prefixes for on-link determination.

   As a result, nodes belonging to the aforementioned prefixes would be
   considered on-link, and packets destined to them would not be relayed
   to a first-hop router.  As a result, the victim nodes (i.e., those
   receiving the crafted Router Advertisements) would perform Neighbor
   Discovery for the intended destination, and when ND failed, the
   packets would be discarded.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network-link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).




Gont, et al.             Expires April 25, 2014                [Page 44]

Internet-Draft           ND Security Assessment             October 2013


   As mentioned in [RFC3756] host implementations could mitigate the
   impact of this attack by requiring the advertised prefixes to be at
   least /64s.

   This would prevent an attacker from affecting a much larger portion
   of the IPv6 address space by e.g. sending a Router Advertisement that
   advertises the prefix ::/0 to be 'on-link'.

   As with other attacks based on Router Advertisement messages, they
   could be mitigated if Layer-2 switches allow Router Advertisements
   messages to be sent only from specific ports.

6.1.7.  Bogus address configuration prefixes

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, illegitimately
   advertising IPv6 prefixes for stateless address auto-configuration
   (SLAAC).  This prefixes could either be bogon prefixes or prefixes
   owned by a remote site.  An attacker could cause victim systems to
   configure IPv6 addresses using prefixes that would cause the
   resulting traffic to be black-holed.

   This would cause the receiving nodes to configure their addresses
   with those bogus prefixes, and as a result, the resulting traffic
   would possibly be filtered by the network (e.g., if network egress-
   filtering is in place).  Even if the outgoing packets were not
   filtered, these victim systems would not receive the return traffic,
   as the return traffic would either be filtered (in the case of bogon
   prefixes) or delivered to the legitimate destination (in the case of
   prefixes owned by some other party).

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network-link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   As with other attacks based on Router Advertisement messages, they
   could be mitigated if Layer-2 switches allow Router Advertisements
   messages to be sent only from specific ports.

6.1.8.  Disabling routers

   An attacker could send crafted Router Advertisements, Neighbor
   Advertisements, or Router Solicitations to cause the receiving nodes
   to remove the impersonated router from the router list.

   In current implementations, if there are no default routers, a packet
   destined to an off-link node will elicit an ICMPv6 Destination



Gont, et al.             Expires April 25, 2014                [Page 45]

Internet-Draft           ND Security Assessment             October 2013


   Unreachable error message.  In the case of legacy implementations
   compliant with [RFC2461], if there are no default routers, the
   Destination Address would be assumed to be 'on-link', and the victim
   would perform Neighbor Discovery for the destination address in the
   hope of delivering the packet on the local link.

   In the case of the Router Advertisements vector, an attacker would
   send unsolicited Router Advertisements with a Preferred Lifetime
   equal to 0 or to some other small value, thus causing the receiving
   nodes to remove the impersonated router from the default router list.

   Alternatively, an attacker could send forged Neighbor Advertisements
   (either solicited or unsolicited) with the Router flag set to 0, thus
   causing the impersonated router to be removed from the default router
   list.

   Receiving nodes would assume the impersonated router has ceased to be
   a router and has changed to functioning only as a host.

   As a third option, an attacker could send a forged Router
   Solicitation message to a router on the local network link, to cause
   the victim to remove the impersonated router from the router list.
   This attack vector is discussed in more detail in Section 4.1.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   Some IPv6 networks employ the 'RA-Guard' mechanism specified in
   [RFC6105] as the first line of defence against RA-based attack
   vectors.  However, However, as discussed in
   [I-D.ietf-v6ops-ra-guard-implementation], some popular RA-Guard
   implementations can be easily circumvented by leveraging IPv6
   extension headers.  [CVE-2011-2395] is a vulnerability advisory about
   this issue.

      [SI6-Toolkit] is a complete complete IPv6 toolkit that can be
      employed to circumvent the aforementioned RA-Guard
      implementations.

   The rest of the attack vectors discussed in this section could
   possibly be mitigated with a more advanced Layer-2 filtering.

6.1.9.  Tampering with 'on-link determination'

   Section 2.1 of [RFC4861] states that a node considers an address to
   be on-link if:



Gont, et al.             Expires April 25, 2014                [Page 46]

Internet-Draft           ND Security Assessment             October 2013


   o  it is covered by one of the link's prefixes (e.g., as indicated by
      the on-link flag in the Prefix Information option), or

   o  a neighbouring router specifies the address as the target of a
      Redirect message, or

   o  a Neighbor Advertisement message is received for the (target)
      address, or

   o  any Neighbor Discovery message is received from the address.

   As a result, some implementations create a Destination Cache entry
   for the Source Address of a Neighbor Discovery message (or for the
   Target Address of a Neighbor Advertisement message) when such a
   message is received, and mark the aforementioned address as 'on-
   link'.

   This means in all traffic meant to the forged address will be
   delivered to the node identified in the corresponding Neighbor Cache
   entry (as the node will be considered to be on-link).  If the
   corresponding Neighbor Cache entry maps the forged address into a
   non-existent or malicious node, all traffic can be black-holed, thus
   leading to a DoS scenario.

   [RFC5942] updates [RFC4861], removing the third and fourth bullets in
   the above list.  This means that receipt of ND messages must not
   result in the Source Address of the ND message or the Target Address
   of a Neighbor Advertisement message to be considered on-link (e.g.,
   by modifying the Prefix List or by marking the corresponding
   Destination Cache entry as 'on-link').

      [CVE-2008-2476] and [US-CERT2008] are vulnerability advisories
      about this issue.

   Some IPv6 networks employ the 'RA-Guard' mechanism specified in
   [RFC6105] as the first line of defence against RA-based attack
   vectors.  However, as discussed in
   [I-D.ietf-v6ops-ra-guard-implementation], some popular RA-Guard
   implementations can be easily circumvented by leveraging IPv6
   extension headers.  [CVE-2011-2395] is a vulnerability advisory about
   this issue.

      [SI6-Toolkit] is a complete complete IPv6 toolkit that can be
      employed to circumvent the aforementioned RA-Guard
      implementations.

   [I-D.ietf-6man-nd-extension-headers] updates [RFC3971] and [RFC4861],
   deprecating the use of fragmentation with Neighbor Discovery, such



Gont, et al.             Expires April 25, 2014                [Page 47]

Internet-Draft           ND Security Assessment             October 2013


   that layer-2 filtering and Neighbor Discovery monitoring become
   feasible.

6.1.10.  Introducing forwarding loops at routers

   As discussed in Section 3.6.2 of this document, if broadcast or
   multicast addresses were allowed in source link-layer address options
   or in target link-layer address options, traffic directed to a victim
   IPv6 address would be sent to such broadcast or multicast IPv6
   address.

   Consider the following network scenario:


             +----------+      +--------+         +----------+
             | Router A |      | Host C |         | Router B |
             +----------+      +--------+         +----------+
                  ||               ||                  ||
              =================================================
                                ||
                                ||
                           +----------+
                           | Attacker |
                           +----------+

          Figure 15: Example network scenario for forwarding loop

   An attacker could poison the neighbor cache of Router A and the
   neighbor cache of Router B, such that the IPv6 address of Host C maps
   to the Ethernet broadcast address (ff:ff:ff:ff:ff:ff).  Afterwards,
   he could send a packet to the Ethernet broadcast address (ff:ff:ff:
   ff:ff:ff), with an IPv6 Destination Address equal to the IPv6 address
   of Host C. Upon receiving the packet, both Router A and Router C
   would decrement the Hop Limit of the packet, and would resend it to
   the Ethernet broadcast address.  As a result, both Router A and
   Router B would now receive two copies of the same packet (one sent by
   Router A, and another sent by Router B).  This would result in a
   'chain reaction' that would only disappear when the Hop Limit of each
   of the packets is decremented to 0.  The total number of packets, for
   a general scenario in which multiple routers are present on the link
   and are subject of the aforementioned neighbor cache poisoning
   attack, and the attacker sends the initial attack packet with an
   arbitrary Hop Limit (possibly 255 to get the maximum amplification
   factor) is:







Gont, et al.             Expires April 25, 2014                [Page 48]

Internet-Draft           ND Security Assessment             October 2013


                                   HopLimit-1
                                      ---
                                      \          x
                         Packets =    /   Routers
                                      ---
                                      x=0

                  Figure 16: Maximum amplification factor

   This equation does not take into account neither the possible ICMPv6
   Redirect messages that each of the Routers could send, nor the
   possible ICMPv6 'time exceeded in transit' error messages that each
   of the routers could possibly send to the Source Address of the
   packet when each of the 'copies' of the original packet is discarded
   as a result of their Hop Limit being decremented to 0.

   As discussed in Section 3.6.2 of this document, neither broadcast nor
   multicast addresses should be allowed in source link-layer address
   and target link-layer address options.  An additional mitigation
   would be for routers to not forward IPv6 packets on the same
   interface if the link-layer destination address of the packet was a
   broadcast or multicast address.

   It is also possible to introduce a forwarding loop at a router by
   poisoning its neighbor cache such that a victim IPv6 address
   (considered to be on-link) maps to one of the attacked router's link-
   layer addresses.  An attacker could poison the neighbor cache of the
   target router as described, and then send a packet to the attacked
   router with the IPv6 Destination Address set to the victim address.
   Upon receipt of the packet, the router would decrement the Hop Limit,
   and 'forward' the packet to its own link-layer address.  This would
   result in a loop, with the target router processing the packet 'Hop
   Limit' times (where 'Hop Limit' is the value used for the Hop Limit
   field of the original packet).

6.1.11.  Tampering with a Neighbor Discovery implementation

   The Neighbor Discovery specification describes conceptual data
   structures such as the Neighbor Cache and the Destination Cache,
   which grow as a result of each entry that is created.  Additionally,
   there are other structures such as the list of configured IPv6
   addresses, the list of Recursive DNS Servers, etc., that also grow
   for each entry that is created in them.

   As discussed throughout Section 5 of this document, an implementation
   should enforce limits on the maximum number of entries in these
   structures.  Failure in enforcing such limits could result in buffer
   overflows or memory exhaustion.



Gont, et al.             Expires April 25, 2014                [Page 49]

Internet-Draft           ND Security Assessment             October 2013


      FreeBSD 9.0 and NetBSD 5.1 fail to enforce limits on the number of
      entries in the IPv6 routing table, on the number of entries in the
      Neighbor Cache, on the number of entries in the Default Router
      List, and on the number of configured IPv6 addresses.  Therefore
      they are vulnerable to multiple Denial of Service attacks.

      Many versions of Windows that support IPv6 fail to enforce limits
      on the number of entries in the IPv6 routing table, on the maximum
      number of configured addresses, and on the number of entries in
      the Neighbor Cache.  Therefore, these structures could be
      exploited for performing a Denial of Service attack.  [Win-Update]
      describes an update has been made available for Windows 7 and
      Windows Server 2008 R2 to limit the number of configured addresses
      and the number of routing table entries on a per-interface basis.

      Linux 2.6.38-10 does enforce a limit on the number of entries in
      the Default router list.  However, this limit itself could be
      leveraged for performing a Denial of Service attack, by causing
      the Default router list to become full of malicious/spurious
      entries before a legitimate entry can be added.  As a result, the
      system would be unable to configure a legitimate default router,
      even if a legitimate Router Advertisement is received at some
      point later.

   An attacker attached to the same network link as the target node can
   stress most of these data structures by sending a large number of the
   appropriate Neighbor Discovery options (e.g., RDNSS or Prefix
   Information options in Router Advertisement messages, etc.) as has
   been shown by e.g.  [CVE-2010-4669].

   Other structures (such as the Neighbor Cache or the Destination
   Cache) can be stressed by sending packets with forged addresses to
   the target node.  For example, an attacker could send any packets
   that would elicit a response from the destination system with forged
   IPv6 Source Address that is assumed to be 'on-link' by the target
   system.  In order for the target node to respond to those packets, it
   would have to create the necessary entries in the Destination Cache
   and in the Neighbor Cache.  If the target implementation does not
   enforce limits on the maximum number of entries in each of those data
   structures, the attack may result in buffer overflows or kernel
   system memory exhaustion.

   It is interesting to note that this attack vector could also be
   exploited by an attacker located in a remote site, unless ingress
   and/or egress filtering are in place.






Gont, et al.             Expires April 25, 2014                [Page 50]

Internet-Draft           ND Security Assessment             October 2013


      [NISCC2006b] discusses ingress and egress filtering.

6.1.12.  Tampering with a Neighbor Discovery router implementation from
         a remote site

   A remote attacker could potentially perform a Denial-of-Service (DoS)
   attack against a router by sending packets to different IPv6
   addresses considered on-link at one of the network links to which the
   target router is attached.  Each of these packets would engage the
   target router in neighbor discovery for each of those addresses,
   probably preventing the router from performing neighbor discovery for
   legitimate packets aimed at existing nodes.

   This problem would be exacerbated if an implementation queues in
   memory those packets that are destined to an IPv6 address for which
   address resolution is being performed.  See Section 5 of this
   document for a thorough description of this issue.

   One important difference between this attack vector and the ones
   described in the previous subsections is that in order for an
   attacker to successfully perform this attack, he does not need to be
   attached to the same network link to which the target router is
   attached.

   A possible mitigation for this attack would be to enforce a limit on
   the maximum number of entries in the Neighbor Cache that are in the
   'INCOMPLETE' state.  This limit should be stricter than the overall
   limit on the maximum number of entries in the Neighbor Cache.

   A Neighbor Cache entry is in the 'INCOMPLETE' state if a Neighbor
   Advertisement message has never been received for the corresponding
   IPv6 address since the entry was created.

   It should be noted that this is an implementation issue rather than a
   protocol-based vulnerability.  However, a number of implementations
   have been found to be vulnerable to this attack.

   It is also worth noting that this attack does not require an attacker
   to forge the IPv6 Source Address of the 'malicious' packets.
   Therefore, mechanisms such as 'ingress filtering' do not provide any
   mitigation for this attack.

   Section 6.1.11 describes another attack vector for stressing the
   Neighbor Cache (and the Destination cache) of both host and router
   implementations.






Gont, et al.             Expires April 25, 2014                [Page 51]

Internet-Draft           ND Security Assessment             October 2013


6.2.  Performance degrading

6.2.1.  Parameter spoofing

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising a
   legitimate default router, but malicious network parameters.

   An attacker could also advertise a small link MTU causing the victim
   nodes to enforce such a small MTU for the corresponding network link.
   This would increase the overhead (headers/data ratio), and possibly
   result in a packet-rate increase (if the same throughput is to be
   maintained).  Additionally, this might also require the use of IPv6
   fragmentation when data are to be transferred across this network
   link.  This is a moderate version of the Denial-of-Service (DoS)
   attack discussed in Section 6.1.5 of this document.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   Some IPv6 networks employ the 'RA-Guard' mechanism specified in
   [RFC6105] as the first line of defence against RA-based attack
   vectors.  However, as discussed in
   [I-D.ietf-v6ops-ra-guard-implementation], some popular RA-Guard
   implementations can be easily circumvented by leveraging IPv6
   extension headers.  [CVE-2011-2395] is a vulnerability advisory about
   this issue.

      [SI6-Toolkit] is a complete complete IPv6 toolkit that can be
      employed to circumvent the aforementioned RA-Guard
      implementations.

6.3.  Traffic hijacking

6.3.1.  Neighbor Cache poisoning

   Neighbor Solicitation and Neighbor Advertisement messages can be
   exploited to maliciously poison the Neighbor Cache of a target node
   such that an IPv6 address maps into the link-layer address of a
   malicious node operated by an attacker.  As a result, once the
   victim's Neighbor Cache is poisoned, the attacker would receive all
   traffic aimed at the victim node.

   This is similar to the Denial-of-Service (DoS) attack described in
   Section 6.1.1 of this document, with the only difference being that
   in this case traffic would be directed to a node operated by the



Gont, et al.             Expires April 25, 2014                [Page 52]

Internet-Draft           ND Security Assessment             October 2013


   attacker, rather than to a non-existent node.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   An attacker could also poison the Neighbor Cache of a target node
   mapping a victim IPv6 address to a multicast or broadcast link-layer
   address, such that he can receive a copy of those packets sent by the
   attacked node to the victim node.  This specific attack vector is
   thoroughly discussed in Section 3.6.2 of this document.

   The same mitigation techniques as described in Section 6.1.1 of this
   document apply to this attack-vector.

6.3.2.  Rogue Router

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising
   his own node as a default router.

   This is similar to the Denial-of-Service (DoS) attack described in
   Section 6.1.4, with the only difference that in this case traffic
   would be directed to a node operated by the attacker, rather than to
   a non-existing node.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   The same mitigation techniques as described in Section 6.1.4 apply to
   this attack vector.

6.3.3.  Bogus on-link prefixes

   An attacker could either send unsolicited Router Advertisements
   and/or illegitimately respond to Router Solicitations, advertising
   bogus prefixes for on-link determination.

   As a result, nodes belonging to the aforementioned prefixes would be
   considered on-link, and packets destined to them would not be relayed
   to a first-hop router, but would instead be delivered on the local
   link.  The victim nodes (i.e., those receiving the crafted Router
   Advertisements) would perform Neighbor Discovery for the intended
   destination, and the attacker could then respond with Neighbor
   Advertisements that advertise the link-layer address of his node, so



Gont, et al.             Expires April 25, 2014                [Page 53]

Internet-Draft           ND Security Assessment             October 2013


   that packets are finally delivered to his malicious node.

   In order for an attacker to successfully perform this attack, he
   would need to be attached to the same network link on which the
   attack is to be launched, or control a node attached to that network
   link (e.g., compromise such a node).

   The same mitigation techniques as described in Section 6.1.6 apply to
   this attack vector.

6.3.4.  Tampering with 'on-link determination'

   This attack is similar to the Denial-of-Service (DoS) attack
   described in Section 6.1.10, with the only difference that for the
   purpose of traffic-hijacking, an attacker would make sure that the
   cached link-layer address of the Neighbor Cache entry corresponding
   to the victim address (the Source Address of the forged Neighbor
   Discovery message or the forged Target Address of the forged Neighbor
   Advertisement message) corresponds to the link-layer address of a
   node operated by the attacker.

   As discussed in Section 6.1.9, [RFC5942] updates [RFC4861], such that
   this attack vector is eliminated.  The same mitigations discussed in
   Section 6.1.9 of this document apply to mitigate this vulnerability.

      [CVE-2008-2476] and [US-CERT2008] are vulnerability advisories
      about this issue.

6.4.  Miscellaneous security issues

6.4.1.  Detecting Sniffing Hosts

   If a system reacts differently depending on whether the network
   interface is in promiscuous mode, this can be leveraged by an
   attacker that is on-link to infer whether the target node is in
   promiscuous mode.  Such a security issue has been found on many
   operating systems, where a packet with a multicast MAC address that
   is not being listened on by that target will be processed only if the
   receiving node is in promiscuous mode (i.e., "sniffing" the network).
   This test can be performed with any packet type, e.g.  Neighbor
   Solicitation or Echo Request.

   [CVE-2010-4562] is one vulnerability advisory about such an issue.








Gont, et al.             Expires April 25, 2014                [Page 54]

Internet-Draft           ND Security Assessment             October 2013


7.  IANA Considerations

   This document has no actions for IANA.
















































Gont, et al.             Expires April 25, 2014                [Page 55]

Internet-Draft           ND Security Assessment             October 2013


8.  Security Considerations

   This entire document is about security vulnerabilities that have been
   found popular Neighbor Discovery implementations, and other potential
   security issues that might be affecting existing implementations.
   This document not only discusses the aforementioned issues, but also
   provides implementation guidance such that these issues can be
   eliminated from the affected implementations and completely avoided
   or mitigated in any new Neighbor Discovery implementations.

   The ultimate goal of this document is to help improve the overall
   maturity of Neighbor Discovery implementations, and to raise
   awareness about current security issues that might affect IPv6
   networks.





































Gont, et al.             Expires April 25, 2014                [Page 56]

Internet-Draft           ND Security Assessment             October 2013


9.  Acknowledgements

   Marc Heuse contributed text, edits, comments, and new vulnerabilities
   that were incorporated into this document.

   The author would like to thank George Kargiotakis, who provided
   valuable comments on earlier versions of this document.

   This document is based on the technical report "Security Assessment
   of the Internet Protocol version 6 (IPv6)" [CPNI-IPv6] authored by
   Fernando Gont on behalf of the UK Centre for the Protection of
   National Infrastructure (CPNI).  The author would like to thank (in
   alphabetical order) Ran Atkinson, Fred Baker, Brian Carpenter, Roque
   Gagliano, Guillermo Gont, Alfred Hoenes, Qing Li, Neil Long, and
   Pekka Savola, for providing valuable feedback on earlier versions of
   such document.  Additionally, the author would like to thank (in
   alphabetical order) Ran Atkinson, Brian Carpenter, Joel M. Halpern,
   Robert Hinden, Pekka Savola, Fred Templin, and Ole Troan, who
   generously answered a number of questions when authoring the
   aforementioned document.































Gont, et al.             Expires April 25, 2014                [Page 57]

Internet-Draft           ND Security Assessment             October 2013


10.  References

10.1.  Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC2460]  Deering, S. and R. Hinden, "Internet Protocol, Version 6
              (IPv6) Specification", RFC 2460, December 1998.

   [RFC2464]  Crawford, M., "Transmission of IPv6 Packets over Ethernet
              Networks", RFC 2464, December 1998.

   [RFC3122]  Conta, A., "Extensions to IPv6 Neighbor Discovery for
              Inverse Discovery Specification", RFC 3122, June 2001.

   [RFC3756]  Nikander, P., Kempf, J., and E. Nordmark, "IPv6 Neighbor
              Discovery (ND) Trust Models and Threats", RFC 3756,
              May 2004.

   [RFC6275]  Perkins, C., Johnson, D., and J. Arkko, "Mobility Support
              in IPv6", RFC 6275, July 2011.

   [RFC3971]  Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure
              Neighbor Discovery (SEND)", RFC 3971, March 2005.

   [RFC4191]  Draves, R. and D. Thaler, "Default Router Preferences and
              More-Specific Routes", RFC 4191, November 2005.

   [RFC4389]  Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
              Proxies (ND Proxy)", RFC 4389, April 2006.

   [RFC4861]  Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
              "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
              September 2007.

   [RFC4862]  Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless
              Address Autoconfiguration", RFC 4862, September 2007.

   [RFC4943]  Roy, S., Durand, A., and J. Paugh, "IPv6 Neighbor
              Discovery On-Link Assumption Considered Harmful",
              RFC 4943, September 2007.

   [RFC6106]  Jeong, J., Park, S., Beloeil, L., and S. Madanapalli,
              "IPv6 Router Advertisement Options for DNS Configuration",
              RFC 6106, November 2010.

   [RFC5798]  Nadas, S., "Virtual Router Redundancy Protocol (VRRP)



Gont, et al.             Expires April 25, 2014                [Page 58]

Internet-Draft           ND Security Assessment             October 2013


              Version 3 for IPv4 and IPv6", RFC 5798, March 2010.

   [RFC5942]  Singh, H., Beebee, W., and E. Nordmark, "IPv6 Subnet
              Model: The Relationship between Links and Subnet
              Prefixes", RFC 5942, July 2010.

   [I-D.ietf-6man-nd-extension-headers]
              Gont, F., "Security Implications of IPv6 Fragmentation
              with IPv6 Neighbor Discovery",
              draft-ietf-6man-nd-extension-headers-05 (work in
              progress), June 2013.

10.2.  Informative References

   [RFC2461]  Narten, T., Nordmark, E., and W. Simpson, "Neighbor
              Discovery for IP Version 6 (IPv6)", RFC 2461,
              December 1998.

   [RFC6104]  Chown, T. and S. Venaas, "Rogue IPv6 Router Advertisement
              Problem Statement", RFC 6104, February 2011.

   [RFC6105]  Levy-Abegnoli, E., Van de Velde, G., Popoviciu, C., and J.
              Mohacsi, "IPv6 Router Advertisement Guard", RFC 6105,
              February 2011.

   [I-D.ietf-v6ops-ra-guard-implementation]
              Gont, F., "Implementation Advice for IPv6 Router
              Advertisement Guard (RA-Guard)",
              draft-ietf-v6ops-ra-guard-implementation-07 (work in
              progress), November 2012.

   [CPNI-IPv6]
              Gont, F., "Security Assessment of the Internet Protocol
              version 6 (IPv6)",  UK Centre for the Protection of
              National Infrastructure, (available on request).

   [CPNI-TCP]
              CPNI, "Security Assessment of the Transmission Control
              Protocol (TCP)", 2009, <http://www.gont.com.ar/papers/
              tn-03-09-security-assessment-TCP.pdf>.

   [Hogg-Vyncke]
              Hogg, S. and E. Vyncke, "IPv6 Security",  Cisco Press; 1
              edition, 2008.

   [Lecigne-Neville-Neil]
              Lecigne, C. and G. Neville-Neil, "Walking through FreeBSD
              IPv6 stack", 2006, <http://clem1.be/gimme/ipv6sec.pdf>.



Gont, et al.             Expires April 25, 2014                [Page 59]

Internet-Draft           ND Security Assessment             October 2013


   [Beck2007]
              Beck, F., Cholez, T., Festor, O., and I. Chrisment,
              "Monitoring the Neighbor Discovery Protocol",  The Second
              International Workshop on IPv6 Today - Technology and
              Deployment - IPv6TD 2007, <http://hal.inria.fr/docs/00/15/
              35/58/PDF/IPv6TD07_beck.pdf>.

   [Beck2007b]
              Beck, F., Festor, O., and I. Chrisment, "IPv6 Neighbor
              Discovery Protocol based OS fingerprinting",  INRIA
              Rapport Technique No 0345, 2007, <http://
              hal.archives-ouvertes.fr/docs/00/18/48/51/PDF/
              RT-0345.pdf>.

   [NDPMon]   "NDPMon - IPv6 Neighbor Discovery Protocol Monitor",
              <http://ndpmon.sourceforge.net/>.

   [arpwatch]
              LBNL/NRG, "arpwatch tool", 2006, <http://ee.lbl.gov/>.

   [NISCC2006b]
              NISCC, "NISCC Technical Note 01/2006: Egress and Ingress
              Filtering", 2006, <http://www.niscc.gov.uk/niscc/docs/
              re-20060420-00294.pdf?lang=en>.

   [vanHauser2006]
              vanHauser, "Attacking the IPv6 Protocol Suite",  EuSecWest
              2006 Conference,
              <http://www.eusecwest.com/esw06/esw06-vanhauser.pdf>.

   [SI6-Toolkit]
              "SI6 Networks' IPv6 toolkit",
              <http://www.si6networks.com/tools/ipv6toolkit>.

   [THC-IPv6]
              "The Hacker's Choice IPv6 Attack Toolkit",
              <http://www.thc.org/thc-ipv6/>.

   [CVE-2012-notyet]
              CVE, "CVE-2012-notyet - entry is upcoming ... to be
              filled", 2012.

   [CVE-2011-2391]
              CVE, "CVE-2011-2391 - IPv6 Neighbor Discovery Protocol
              (NDP) implementations do not limit the rate of Neighbor
              Discovery messages processed", 2011.

   [CVE-2008-2476]



Gont, et al.             Expires April 25, 2014                [Page 60]

Internet-Draft           ND Security Assessment             October 2013


              CVE, "CVE-2008-2476 - IPv6 Neighbor Discovery Protocol
              (NDP) implementations do not validate the origin of
              Neighbor Discovery messages", 2008.

   [CVE-2010-4669]
              CVE, "CVE-2010-4669 - Neighbor Discovery (ND) protocol
              implementation in the IPv6 stack in Microsoft Windows
              allows attackers to cause a denial of service (CPU
              consumption and system hang) by sending many Router
              Advertisement (RA) messages with different source
              addresses", 2010.

   [CVE-2011-2395]
              CVE, "CVE-2011-2395 - Neighbor Discovery (ND) protocol
              implementation in Cisco IOS on unspecified switches allows
              attackers to bypass the Router Advertisement Guarding
              functionality via a fragmented IPv6 packets", 2011.

   [CVE-2010-4562]
              CVE, "CVE-2010-4562 - Microsoft Windows, when using IPv6,
              allows remote attackers to determine whether a host is
              sniffing the network by sending an ICMPv6 Echo Request to
              a multicast address and determining whether an Echo Reply
              is sent", 2010.

   [US-CERT2008]
              US-CERT, "US-CERT Vulnerability Note VU#472363: IPv6
              implementations insecurely update Forwarding Information
              Base", 2008.

   [Win-Update]
              Microsoft, "An IPv6 readiness update is available for
              Windows 7 and for Windows Server 2008 R2", 2012.


















Gont, et al.             Expires April 25, 2014                [Page 61]

Internet-Draft           ND Security Assessment             October 2013


Authors' Addresses

   Fernando Gont
   SI6 Networks / UTN-FRH
   Evaristo Carriego 2644
   Haedo, Provincia de Buenos Aires  1706
   Argentina

   Phone: +54 11 4650 8472
   Email: fgont@si6networks.com
   URI:   http://www.si6networks.com


   Ronald P. Bonica
   Juniper Networks
   2251 Corporate Park Drive
   Herndon, VA  20171
   US

   Phone: 571 250 5819
   Email: rbonica@juniper.net


   Will Liu
   Huawei Technologies
   Bantian, Longgang District
   Shenzhen  518129
   P.R. China

   Email: liushucheng@huawei.com





















Gont, et al.             Expires April 25, 2014                [Page 62]


Html markup produced by rfcmarkup 1.108, available from http://tools.ietf.org/tools/rfcmarkup/