RFC 6018

Internet Engineering Task Force (IETF)                          F. Baker
Request for Comments: 6018                                 Cisco Systems
Category: Informational                                        W. Harrop
ISSN: 2070-1721                                              G. Armitage
                                      Swinburne University of Technology
                                                          September 2010

                         IPv4 and IPv6 Greynets


   This note discusses a feature to support building Greynets for IPv4
   and IPv6.

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

Copyright Notice

   Copyright (c) 2010 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.

Baker, et al.                 Informational                     [Page 1]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . 2
     1.1.  History and Experience  . . . . . . . . . . . . . . . . . . 3
   2.  Deploying Greynets  . . . . . . . . . . . . . . . . . . . . . . 4
     2.1.  Deployment Using Routing - Darknets . . . . . . . . . . . . 4
     2.2.  Deployment Using Sparse Address Space - Greynets  . . . . . 4
     2.3.  Other Filters . . . . . . . . . . . . . . . . . . . . . . . 6
   3.  Implications for Router Design  . . . . . . . . . . . . . . . . 6
   4.  Security Considerations . . . . . . . . . . . . . . . . . . . . 6
   5.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . . . 7
   6.  References  . . . . . . . . . . . . . . . . . . . . . . . . . . 8
     6.1.  Normative References  . . . . . . . . . . . . . . . . . . . 8
     6.2.  Informative References  . . . . . . . . . . . . . . . . . . 8

1.  Introduction

   Darknets, also called "Network Telescopes" among other things, have
   been deployed by several organizations (including CAIDA, Team Cymru,
   and the University of Michigan) to look at traffic directed to
   addresses in blocks that are not in actual use.  Such traffic becomes
   visible by either direct capture (it is routed to a collector) or by
   virtue of its backscatter (its resulting in ICMP traffic or
   transport-layer resets).

   Darknets, of course, have two problems.  As their address spaces
   become known, attackers stop probing them, so they are less
   effective.  Also, the administrators of those prefixes are pressured
   by Regional Internet Registry (RIR) policy and business requirements
   to deploy them in active networks.

   [Harrop] defines a 'Greynet' by extension, in these words:

      Darknets are often proposed to monitor for anomalous, externally
      sourced traffic, and require large, contiguous blocks of unused IP
      addresses - not always feasible for enterprise network operators.
      We introduce and evaluate the Greynet - a region of IP address
      space that is sparsely populated with "darknet" addresses
      interspersed with active (or "lit") IP addresses.  Based on a
      small sample of traffic collected within a university campus
      network we saw that relatively sparse greynets can achieve useful
      levels of network scan detection.

   In other words, instead of setting aside prefixes that an attacker
   might attempt to probe and in so doing court discovery, Harrop
   proposed that individual (or small groups of adjacent) addresses in
   subnets be set aside for the purpose, using different host
   identifiers in each subnet to make it more difficult for an address

Baker, et al.                 Informational                     [Page 2]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   scan to detect them.  The concept has value in the sense that it is
   harder to map the addresses or prefixes out of an attacker's search
   pattern, as their presence is more obscure.  Harrop's research was
   carried out using IPv4 [RFC0791] and yielded interesting information.

1.1.  History and Experience

   The research supporting this proposal includes two prototypes, one
   with IPv4 [RFC0791] and one with IPv6 [RFC2460].  Both have
   limitations, being research experiments as opposed to deployment of a
   finished product.

   The original research was done by Warren Harrop and documented in
   [Harrop].  This was IPv4-only.  His premise was that one would put a
   virtual or physical machine on a LAN that one was not otherwise
   using, and use it to identify scans of various kinds.  As reported in
   his paper, the concept worked effectively in a prototype deployment
   at the Centre for Advanced Internet Architectures (CAIA), Swinburne
   University of Technology.  The basic reason was that there was a
   reasonable expectation on the part of a potential attacker that a
   given address might be represented, and there was no pattern that
   would enable the attacker to predict which addresses were being used
   in this way.  CAIA developed and released a prototype FreeBSD-based
   Greynet system in 2008 built around this premise [Armitage].

   Baker's addition to his concept started from the router, the idea
   that the router would be highly likely to encounter any such scan if
   it came from off-LAN, and the fact that the router would have to use
   Address Resolution Protocol (ARP) or Neighbor Discovery (ND) to
   identify -- or fail to identify -- the machine in question.  In
   effect, any address that is not currently instantiated in the subnet
   acts as a Greynet trigger address.  This clearly also works for any
   system that would implement ARP or ND, but the router is an obvious
   focal point in any subnet.

   Tim Chown, of the School of Electronics and Computer Science,
   University of Southampton, offered privately to do some research on
   it, and had Owen Stephens do a Linux prototype in spring 2010.  They
   demonstrated that the technology was straightforward to implement and
   in fact worked in a prototype IPv6 implementation.

   The question that remains with IPv6 address scanning is the
   likelihood that the attack would occur at all.  Chown originally
   argued in [RFC5157] that address scans were impossible due to the
   sheer number of possibilities.  However, in September 2010 a report
   was made to NANOG of an IPv6 address scan.  Additionally, there are
   ways to limit the field; for example, one can observe that a company
   buys a certain kind of machine or network interface card (NIC), and

Baker, et al.                 Informational                     [Page 3]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   therefore its probable EUI-64 addresses are limited to a much smaller
   range than 2^64 -- more like 2^24 addresses on a given subnet -- or
   one can observe DNS, SMTP envelopes, Extensible Messaging and
   Presence Protocol (XMPP) messages, FTP, HTTP, etc., that carry IP
   addresses in other ways.  Such attacks can be limited by the use of
   Privacy Addresses [RFC4941], which periodically change, rendering
   historical information less useful, but the fact is that such
   analytic methods exist.

2.  Deploying Greynets

   Corporate IT departments and other network operators frequently run
   collectors or other kinds of sensors.  A collector is a computer
   system on the Internet that is expressly set up to attract and "trap"
   nefarious attempts to penetrate computer systems.  Such systems may
   simply record the attempt or the datagram that initiated the attempt
   (darknets/Greynets), or they may act as a decoy, luring in potential
   attacks in order to study their activities and study their methods

   To accomplish this, we separate nefarious traffic from that which is
   likely normal and important, studying one and facilitating the other.

2.1.  Deployment Using Routing - Darknets

   One obvious way to isolate and identify nefarious traffic is to
   realize that it is sent to a prefix or address that is not
   instantiated.  If a campus uses an IPv4 /24 prefix or an IPv6 /56
   prefix but contains less than 100 actual subnets, for example, we
   might use only odd numbered subnets (128 of the 256 available in that
   prefix), and not quite all of those.  Knowing that the active
   prefixes are more specific and therefore attract appropriate traffic,
   we might also advertise the default prefix from the collector,
   attracting traffic directed to the uninstantiated prefixes in that
   routing domain.

   A second question involves mimicking a host under attack; the
   collector may simply record this uninvited traffic, or may reply as a
   honeypot system.

2.2.  Deployment Using Sparse Address Space - Greynets

   IPv4 subnets usually have some unallocated space in them, if only
   because Classless Inter-Domain Routing (CIDR) allocates O(2^n)
   addresses to an IP subnet and there are not exactly that many systems

Baker, et al.                 Informational                     [Page 4]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   Similarly, with active IPv6 prefixes, even a very large switched LAN
   is likely to use a small fraction of the available addresses.  This
   is by design, as discussed in Section 2.5.1 of [RFC4291].  If the
   addresses are distributed reasonably randomly among the possible
   values, the likelihood of an attacker guessing what addresses are in
   actual use is limited.  This gives us an opportunity with respect to
   unused addresses within an IP prefix.

   Routers use IPv4 ARP [RFC0826] and IPv6 Neighbor Discovery [RFC4861]
   to determine the MAC (Media Access Control) address of a neighbor to
   which a datagram needs to be sent.  Both specifications intend that
   when a datagram arrives at a router that serves the target prefix,
   but that doesn't know the MAC address of the intended destination, it

   o  Enqueue the datagram,

   o  Emit a Neighbor Solicitation or ARP Request,

   o  Await a Neighbor Advertisement or ARP Response, and

   o  On receipt, dequeue and forward the datagram.

   Once the host's MAC address is in the router's tables (and in so
   doing the address proven valid), the matter is not an issue.

   In [Harrop], the Greynet is described as being instantiated on an
   end-host that replies to ARP Requests for all 'dark' IP addresses.
   However, a small modification to router behavior can augment this
   model.  As well as queuing or dropping a datagram that has triggered
   an ARP Request or Neighbor Solicitation, the router forwards a copy
   of this datagram over an independent link to the Greynet's analytic
   equipment.  This independent link may be a different physical
   interface, a circuit, VLAN, tunnel, UDP, or other encapsulation, or
   in fact any place such a datagram could be handled.  Depending on the
   requirements of the receiving collector, one could also imagine
   summarizing information in a form similar to IP Flow Information
   Export (IPFIX) [RFC5101] [RFC5610].

   The analytic equipment will now receive two types of datagrams.  Of
   most interest will be those destined for 'dark' IP addresses.  Of
   less interest will be the irregular case where a datagram arrives for
   a legitimate local neighbor who has, for some temporary reason, no
   MAC address in the router's tables.  Datagrams arriving for an IP
   destination for which an ARP reply (or Neighbor Advertisement) has
   not yet received might also be forwarded to the analytical equipment
   over the independent link -- or might not, if they are considered to
   be unlikely to provide new analytic information.

Baker, et al.                 Informational                     [Page 5]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   Analytic equipment, depending on the router to recognize 'dark' IP
   addresses in this manner, can easily track arrival patterns of
   datagrams destined to unused parts of the network.  It may also
   optionally choose to respond to such datagrams, acting as a honeypot
   to elicit further datagrams from the remote source.

   If the collector replies directly, the attacker may be able to
   identify the fact through information in or about the datagram -
   datagrams sent to the same IP subnet may come back with different TTL
   values, for example.  Hence, it may be advisable for the collector to
   send the reply back through the tunnel and therefore as if from the
   same IP subnet.  Naturally, the collector in this scenario should not
   respond to datagrams destined for 'lit' IP addresses -- the intended
   destination will eventually respond to the router's ARP or Neighbor
   Solicitation anyway.

   One implication of this model is that distributed denial-of-service
   (DDoS) attacks terminate on router subnets within a network, as
   opposed to stopping on inter-router links.

2.3.  Other Filters

   An obvious extension of the concept would include traffic identified
   by other filters as appropriate to send to the collector.  For
   example, one might configure the system to forward traffic that fail
   a unicast Reverse Path Forwarding (uRPF) check [RFC2827] to the
   collector via the same tunnel.

3.  Implications for Router Design

   The implication for router design applies to the IPv4 ARP and IPv6
   Neighbor Discovery algorithms.  It might be interesting to provide,
   under configuration control, the ability to forward to an analytic
   system the arriving datagrams that trigger an ARP Request or Neighbor
   Solicit, and then fail to receive the intended response, to an
   interface, circuit, VLAN, or tunnel.

4.  Security Considerations

   This note describes a tool for managing IPv4 and IPv6 network
   security.  Like any tool, it has limitations and possible attacks.
   If discarding traffic under overload is a good thing, then holding
   and subsequently forwarding the traffic instead places a potential
   load on the network and the router in question, and as such
   represents a possible attack.  Such an attack has obvious
   mitigations, however; one simply selects (in a manner the operator
   deems appropriate) a subset of the traffic to forward and discards
   the rest.  In addition, this attack is not new; it is only changed in

Baker, et al.                 Informational                     [Page 6]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   character.  A stream that would instantiate the attack today results
   in a load of ARP or Neighbor Solicit messages that all listening
   hosts must intelligently discard.  The new attack additionally
   consumes bandwidth that is presumably set aside specifically for that

   The question of exactly what subset of traffic is interesting and
   economical to forward is intentionally left open.  Key questions in
   algorithm design include what can be learned from a given sample (Are
   bursts happening?  If so, with what data?), what the impact on the
   router and other equipment in question is, how that might be
   mitigated, etc.  Possible selection algorithms dependent only on
   state and algorithms typically available in a router include:

   o  Select all datagrams that trigger an ARP Request or Neighbor

   o  Select the subset of those that are not responded to within some
      stated interval and are therefore likely dark.

   o  Select the subset of those that are new; if the address is
      currently being solicited, forwarding redundant data may not be

   o  Select all datagrams up to some rate.

   o  Select all datagrams matching (or not matching) a specified filter

5.  Acknowledgements

   Algorithms for learning about Internet attack behavior by observing
   backscatter traffic have been used by CAIDA, University of Michigan,
   Team Cymru, and others.  Harrop extended them in his research.  This
   formulation of the notion originated in a discussion among the
   authors in 2005.  This note grew out of a conversation with Paul
   Vixie and Rhette Marsh on Internet traffic sensors; they also made
   useful comments on it.  Albert Manfredi commented on the distinction
   between a LAN (as defined by IEEE 802) and an IP subnet.

   Tim Chown [RFC5157] has observed that, at least at the time of
   writing that RFC, address scanning attacks in IPv6 have not been
   reported in the wild.  However, as mentioned in Section 1.1 above, a
   (partial) scanning attack was recently reported on the NANOG mailing
   list.  Rhette Marsh has suggested the structure of such an attack,
   however, and Fred Baker has suggested approaches based on addressing

Baker, et al.                 Informational                     [Page 7]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   information exchanged by applications.  Hence, we believe that such
   issues may be relevant to IPv6 in the future, when IPv6 is a more
   interesting target.

   Tim Chown and Owen Stephens tested the proposal, and made useful
   comments that have been incorporated in this text.  His fundamental
   comment was, however, that "it works".

6.  References

6.1.  Normative References

   [Harrop]   Harrop, W. and G. Armitage, "Greynets: a definition and
              evaluation of sparsely populated darknets", IEEE LCN IEEE
              30th Conference on Local Computer Networks, 2005.

   [RFC0791]  Postel, J., "Internet Protocol", STD 5, RFC 791,
              September 1981.

   [RFC0826]  Plummer, D., "Ethernet Address Resolution Protocol: Or
              converting network protocol addresses to 48.bit Ethernet
              address for transmission on Ethernet hardware", STD 37,
              RFC 826, November 1982.

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

   [RFC4291]  Hinden, R. and S. Deering, "IP Version 6 Addressing
              Architecture", RFC 4291, February 2006.

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

   [RFC4941]  Narten, T., Draves, R., and S. Krishnan, "Privacy
              Extensions for Stateless Address Autoconfiguration in
              IPv6", RFC 4941, September 2007.

6.2.  Informative References

   [Armitage] Armitage, G., Harrop, W., Heyde, A., Parry, L., "Greynets:
              Passive Detection of Unsolicited Network Scans in Small
              ISP and Enterprise networks", CAIA, Swinburne University
              of Technology, December 2008,

Baker, et al.                 Informational                     [Page 8]

RFC 6018                 IPv4 and IPv6 Greynets           September 2010

   [RFC2827]  Ferguson, P. and D. Senie, "Network Ingress Filtering:
              Defeating Denial of Service Attacks which employ IP Source
              Address Spoofing", BCP 38, RFC 2827, May 2000.

   [RFC5101]  Claise, B., "Specification of the IP Flow Information
              Export (IPFIX) Protocol for the Exchange of IP Traffic
              Flow Information", RFC 5101, January 2008.

   [RFC5157]  Chown, T., "IPv6 Implications for Network Scanning",
              RFC 5157, March 2008.

   [RFC5610]  Boschi, E., Trammell, B., Mark, L., and T. Zseby,
              "Exporting Type Information for IP Flow Information Export
              (IPFIX) Information Elements", RFC 5610, July 2009.

Authors' Addresses

   Fred Baker
   Cisco Systems
   Santa Barbara, California  93117

   EMail: fred@cisco.com

   Warren Harrop
   Centre for Advanced Internet Architectures
   Swinburne University of Technology
   PO Box 218
   John Street, Hawthorn,
   Victoria, 3122

   EMail: wazz@bud.cc.swin.edu.au

   Grenville Armitage
   Centre for Advanced Internet Architectures
   Swinburne University of Technology
   PO Box 218
   John Street, Hawthorn,
   Victoria, 3122

   EMail: garmitage@swin.edu.au

Baker, et al.                 Informational                     [Page 9]