RFC 6666 INFORMATIONAL IPV6 · INFRASTRUCTURE · OPERATIONS

IPv6 Discard Prefix: A Discard Prefix for IPv6

A reserved IPv6 prefix for filtering and discarding unwanted traffic.

On rfc-editor.org ↗
PUBLISHED
August 2012
AUTHORS
N. Hilliard, D. Freedman
STREAM
IETF
DISCARD PREFIX
100::/64
PREFIX LENGTH
64 bits
ADDRESS RANGE
100::0 to 100::ffff:ffff:ffff:ffff
PUBLICATION DATE
August 2012
ALLOCATION
IANA IPv6 Special Purpose Address Registry
DOI
10.17487/RFC6666

RFC text

Contents
Original text · RFC Editor View on rfc-editor.org ↗
Internet Engineering Task Force (IETF)                       N. Hilliard
Request for Comments: 6666                                          INEX
Category: Informational                                      D. Freedman
ISSN: 2070-1721                                                 Claranet
                                                             August 2012


                       A Discard Prefix for IPv6

Abstract

   Remote triggered black hole filtering describes a method of
   mitigating the effects of denial-of-service attacks by selectively
   discarding traffic based on source or destination address.  Remote
   triggered black hole routing describes a method of selectively re-
   routing traffic into a sinkhole router (for further analysis) based
   on destination address.  This document updates the "IPv6 Special
   Purpose Address Registry" by explaining why a unique IPv6 prefix
   should be formally assigned by IANA for the purpose of facilitating
   IPv6 remote triggered black hole filtering and routing.

Status of This Memo

   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Not all documents
   approved by the IESG are a candidate for any level of Internet
   Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc6666.

Copyright Notice

   Copyright (c) 2012 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   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. Introduction ....................................................2
      1.1. Notational Conventions .....................................3
   2. A Discard Prefix for IPv6 .......................................3
   3. Operational Implications ........................................4
   4. IANA Considerations .............................................4
   5. Security Considerations .........................................4
   6. References ......................................................5
      6.1. Normative References .......................................5
      6.2. Informative References .....................................5

1.  Introduction

   Remote Triggered Black Hole (RTBH) filtering describes a class of
   methods of blocking IP traffic either from a specific source
   ([RFC5635]) or to a specific destination ([RFC3882]) on a network.
   RTBH routing describes a class of methods of re-routing IP traffic
   destined to the attacked/targeted host to a special path (tunnel)
   where a sniffer could capture the traffic for analysis.  Both of
   these methods operate by setting the next-hop address of an IP packet
   with a specified source or destination address to be a unicast prefix
   that is connected locally or remotely to a router's discard, null, or
   tunnel interface.  Typically, reachability information for this
   prefix is propagated throughout an autonomous system using a dynamic
   routing protocol such as BGP ([RFC3882]).  By deploying RTBH systems
   across a network, traffic to or from specific destinations may be
   selectively black-holed or re-routed to a sinkhole device in a manner
   that is efficient, scalable, and straightforward to implement.

   On some networks, operators configure RTBH installations using
   [RFC1918] address space or the address blocks reserved for
   documentation in [RFC5737].  This approach is inadequate because RTBH
   configurations are not documentation, but rather operationally
   important features of many public-facing production networks.
   Furthermore, [RFC3849] specifies that the IPv6 documentation prefix
   should be filtered in both local and public contexts.  On this basis,
   it is suggested that both private network address blocks and the
   documentation prefixes described in [RFC5737] are inappropriate for
   RTBH configurations and that a dedicated IPv6 prefix should be
   assigned instead.

   This document updates the "IPv6 Special Purpose Address Registry"
   [IANA-IPV6REG].

1.1.  Notational Conventions

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC2119].

2.  A Discard Prefix for IPv6

   For the purposes of implementing an IPv6 RTBH configuration, a
   unicast address block is required.  There are currently no IPv6
   unicast address blocks that are specifically nominated for the
   purposes of implementing such RTBH systems.

   While it could be argued that there are other addresses and address
   prefixes that could be used for this purpose (e.g., documentation
   prefixes, private address space), or that an operator could assign an
   address block from their own address space for this purpose, there is
   currently no operational clarity on what address block would be
   appropriate or inappropriate to use for this purpose.  By assigning a
   globally unique discard prefix for IPv6, the IETF will introduce good
   practice for the implementation of IPv6 RTBH configurations and will
   facilitate operational clarity by allowing operators to implement
   consistent and deterministic inter-domain prefix and traffic
   filtering policies for black-holed traffic.

   As [RFC3882] and [RFC5635] describe situations where more than one
   discard address may be used for implementing multiple RTBH scenarios,
   a single address is not sufficient to cover all likely RTBH
   situations.  Consequently, an address block is required.

3.  Operational Implications

   This assignment MAY be carried in a dynamic routing protocol within
   an autonomous system.  The assignment SHOULD NOT be announced to or
   accepted from third-party autonomous systems, and IPv6 traffic with a
   destination address within this prefix SHOULD NOT be forwarded to or
   accepted from third-party autonomous systems.  If the prefix or a
   subnet of the prefix is inadvertently announced to or accepted from a
   third-party autonomous system, this may cause excessive volumes of
   traffic to pass unintentionally between the two networks, which would
   aggravate the effect of a denial-of-service attack.

   On networks that implement IPv6 remote triggered black holes, some or
   all of this network block MAY be configured with a next-hop
   destination of a discard or null interface on any or all IPv6 routers
   within the autonomous system.

4.  IANA Considerations

   Per this document, IANA has recorded the allocation of the IPv6
   address prefix 0100::/64 as a Discard-Only Prefix in the "Internet
   Protocol Version 6 Address Space" and added the prefix to the "IANA
   IPv6 Special Purpose Address Registry" [IANA-IPV6REG].  No end party
   has been assigned to this prefix.  The prefix has been allocated from
   ::/3.

5.  Security Considerations

   As the prefix specified in this document ought not normally be
   transmitted or accepted over inter-domain BGP sessions for the
   reasons described in Section 3, it is usually appropriate to include
   this prefix in inter-domain BGP prefix filters [RFC3704] or otherwise
   ensure the prefix is neither transmitted to nor accepted from a
   third-party autonomous system.

6.  References

6.1.  Normative References

   [IANA-IPV6REG]
              Internet Assigned Numbers Authority, "IPv6 Special Purpose
              Address Registry", 2012, <http://www.iana.org/assignments/
              iana-ipv6-special-registry>.

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

   [RFC3882]  Turk, D., "Configuring BGP to Block Denial-of-Service
              Attacks", RFC 3882, September 2004.

   [RFC5635]  Kumari, W. and D. McPherson, "Remote Triggered Black Hole
              Filtering with Unicast Reverse Path Forwarding (uRPF)",
              RFC 5635, August 2009.

6.2.  Informative References

   [RFC1918]  Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot, G.,
              and E. Lear, "Address Allocation for Private Internets",
              BCP 5, RFC 1918, February 1996.

   [RFC3704]  Baker, F. and P. Savola, "Ingress Filtering for Multihomed
              Networks", BCP 84, RFC 3704, March 2004.

   [RFC3849]  Huston, G., Lord, A., and P. Smith, "IPv6 Address Prefix
              Reserved for Documentation", RFC 3849, July 2004.

   [RFC5737]  Arkko, J., Cotton, M., and L. Vegoda, "IPv4 Address Blocks
              Reserved for Documentation", RFC 5737, January 2010.

Authors' Addresses

   Nick Hilliard
   INEX
   4027 Kingswood Road
   Dublin  24
   IE

   EMail: nick@inex.ie


   David Freedman
   Claranet
   21 Southampton Row, Holborn
   London  WC1B 5HA
   UK

   EMail: david.freedman@uk.clara.net

AI ELI5

On this page
PROBLEMNetworks use black hole filtering to drop attack traffic based on destination address, but were using unassigned or documentation address space for this, creating ambiguity about what prefixes should be reserved.
MECHANISMIANA assigns a dedicated IPv6 prefix, 100::/64, as a Discard-Only Prefix specifically for remote triggered black hole filtering and routing.
OPERATIONAL CLARITYNetwork operators can now use a globally consistent prefix for discard filtering, avoiding conflicts with documentation or private address ranges.

§1Remote Triggered Black Hole Filtering

Remote triggered black hole (RTBH) filtering is a method of defending against denial-of-service attacks by selectively discarding traffic destined to a target address. When an attack is detected, routers are configured to direct traffic to that address into a discard interface or tunnel, preventing it from reaching the target.

Earlier, networks implemented RTBH systems by routing attack traffic to private address space (RFC 1918) or to documentation prefixes (RFC 5737). However, these approaches create operational confusion: documentation prefixes should be filtered in both local and public contexts, and private space is meant for private networks, not production defense infrastructure. There was no globally recognized standard prefix for this critical operational task.

When the internet is under attack, routers can automatically send the bad traffic down a special dead-end address so it does not hurt the real servers. This RFC gives everyone an official dead-end address.

§2The Discard Prefix

The RFC specifies that IANA assign the IPv6 prefix 100::/64 (written as 0100::/64 in full IPv6 notation) as a dedicated Discard-Only Prefix. The /64 means the prefix covers a block of addresses from 100::0 to 100::ffff:ffff:ffff:ffff.

This size was chosen to allow multiple RTBH scenarios: as noted in earlier RFCs on RTBH, operators may need more than one discard address to implement different filtering policies or to separate different types of attacks. A full /64 block provides this flexibility.

The official dead-end address range is 100::/64, which is like a whole neighborhood of wrong addresses. It has enough room for many different discard scenarios.

§3Deployment and Filtering

Operators may announce this prefix within their autonomous system using dynamic routing protocols like BGP. However, the prefix should not be announced to or accepted from third-party autonomous systems. If the prefix leaks into inter-domain BGP, traffic intended for discard might reach another network unintentionally, potentially amplifying a denial-of-service attack.

Within a network implementing IPv6 RTBH, routers can be configured with this prefix pointing to a discard or null interface, meaning packets destined to any address in 100::/64 are dropped immediately. BGP prefix filters should be configured to ensure this prefix does not escape the network boundary.

Each network uses the discard prefix only on the inside, like a real dead-end street in your own town. You do not advertise it to other networks, so traffic never escapes.

§4Standards Adoption and Impact

This RFC formalizes the allocation of a globally recognized discard prefix, ending the era of operators improvising with documentation or private address space. The specification resolves ambiguity about what prefix to use for RTBH implementations, making it possible for operators to deploy consistent defense mechanisms across their networks and coordinate with peers.

The discard prefix has been recorded in the IANA IPv6 Special Purpose Address Registry, making it an official part of the IPv6 address architecture. The key operational requirement is that this prefix be isolated within networks where RTBH is deployed: it must not leak into the global routing system, or inter-domain traffic to the discard prefix could cause unintended consequences. Proper BGP filtering and internal routing configuration are essential for safe deployment.

By creating an official dead-end address, the RFC lets everyone use the same one, and networks can coordinate to keep it from escaping to the internet.

§5References