RFC 6748






Internet Research Task Force (IRTF)                          RJ Atkinson
Request for Comments: 6748                                    Consultant
Category: Experimental                                         SN Bhatti
ISSN: 2070-1721                                            U. St Andrews
                                                           November 2012


             Optional Advanced Deployment Scenarios for the
               Identifier-Locator Network Protocol (ILNP)

Abstract



   This document provides an Architectural description and the Concept
   of Operations of some optional advanced deployment scenarios for the
   Identifier-Locator Network Protocol (ILNP), which is an evolutionary
   enhancement to IP.  None of the functions described here is required
   for the use or deployment of ILNP.  Instead, it offers descriptions
   of engineering and deployment options that might provide either
   enhanced capability or convenience in administration or management of
   ILNP-based systems.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for examination, experimental implementation, and
   evaluation.

   This document defines an Experimental Protocol for the Internet
   community.  This document is a product of the Internet Research Task
   Force (IRTF).  The IRTF publishes the results of Internet-related
   research and development activities.  These results might not be
   suitable for deployment.  This RFC represents the individual
   opinion(s) of one or more members of the Routing Research Group of
   the Internet Research Task Force (IRTF).  Documents approved for
   publication by the IRSG are not 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/rfc6748.











Atkinson & Bhatti             Experimental                      [Page 1]

RFC 6748                        ILNP ADV                   November 2012


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.

   This document may not be modified, and derivative works of it may not
   be created, except to format it for publication as an RFC or to
   translate it into languages other than English.




































Atkinson & Bhatti             Experimental                      [Page 2]

RFC 6748                        ILNP ADV                   November 2012


Table of Contents



   1. Introduction ....................................................4
      1.1. Document Roadmap ...........................................5
      1.2. Terminology ................................................6
   2. Localised Numbering .............................................6
      2.1. Localised Locators .........................................7
      2.2. Mixed Local/Global Numbering ...............................9
      2.3. Dealing with Internal Subnets with Locator Rewriting .......9
      2.4. Localised Name Resolution with DNS ........................11
      2.5. Use of mDNS ...............................................13
      2.6. Site Network Name in DNS ..................................13
      2.7. Site Interior Topology Obfuscation ........................14
      2.8. Other SBR Considerations ..................................14
   3. An Alternative for Site Multihoming ............................16
      3.1. Site Multihoming (S-MH) Connectivity Using an SBR .........16
      3.2. Dealing with Link/Connectivity Changes ....................17
      3.3. SBR Updates to DNS ........................................18
      3.4. DNS TTL Values for L32 and L64 Records ....................18
      3.5. Multiple SBRs .............................................19
   4. An Alternative for Site (Network) Mobility .....................20
      4.1. Site (Network) Mobility ...................................20
      4.2. SBR Updates to DNS ........................................22
      4.3. DNS TTL Values for L32 and L64 Records ....................22
   5. Traffic Engineering Options ....................................22
      5.1. Load Balancing ............................................23
      5.2. Control of Egress Traffic Paths ...........................24
   6. ILNP in Datacentres ............................................26
      6.1. Virtual Image Mobility within a Single Datacentre .........27
      6.2. Virtual Image Mobility between Datacentres - Invisible ....28
      6.3. Virtual Image Mobility between Datacentres - Visible ......29
      6.4. ILNP Capability in the Remote Host for VM Image Mobility ..29
   7. Location Privacy ...............................................30
      7.1. Locator Rewriting Relay (LRR) .............................30
      7.2. Options for Installing LRR Packet Forwarding State ........31
   8. Identity Privacy ...............................................32
   9. Security Considerations ........................................32
   10. References ....................................................33
      10.1. Normative References .....................................33
      10.2. Informative References ...................................34
   11. Acknowledgements ..............................................37










Atkinson & Bhatti             Experimental                      [Page 3]

RFC 6748                        ILNP ADV                   November 2012


1.  Introduction



   This document is part of the ILNP document set, which has had
   extensive review within the IRTF Routing RG.  ILNP is one of the
   recommendations made by the RG Chairs.  Separately, various refereed
   research papers on ILNP have also been published during this decade.
   So, the ideas contained herein have had much broader review than the
   IRTF Routing RG.  The views in this document were considered
   controversial by the Routing RG, but the RG reached a consensus that
   the document still should be published.  The Routing RG has had
   remarkably little consensus on anything, so virtually all Routing RG
   outputs are considered controversial.

   At present, the Internet research and development community is
   exploring various approaches to evolving the Internet Architecture to
   solve a variety of issues including, but not limited to, scalability
   of inter-domain routing [RFC4984].  A wide range of other issues
   (e.g., site multihoming, node multihoming, site/subnet mobility, node
   mobility) are also active concerns at present.  Several different
   classes of evolution are being considered by the Internet research
   and development community.  One class is often called "Map and
   Encapsulate", where traffic would be mapped and then tunnelled
   through the inter-domain core of the Internet.  Another class being
   considered is sometimes known as "Identifier/Locator Split".  This
   document relates to a proposal that is in the latter class of
   evolutionary approaches.

   ILNP is, in essence, an end-to-end architecture: the functions
   required for ILNP are implemented in, and controlled by, only those
   end-systems that wish to use ILNP, as described in [RFC6740].  Other
   nodes, such as Site Border Routers (SBRs) need only support IP to
   allow operation of ILNP, e.g., an SBR should support IPv6 in order to
   enable end-systems to operate ILNPv6 within the site network for
   which an SBR provides a service [RFC6741].

   However, some features of ILNP could be optimised, from an
   engineering perspective, by the use of an intermediate system (a
   router, security gateway or "middlebox") that modifies (rewrites)
   Locator values of transit ILNP packets.  It would also perform other
   control functions for an entire site, as an administrative
   convenience, such as providing a centralised point of management for
   a site.  For example, an SBR might manipulate the topological
   presence of the packet, providing an elegant solution to the
   provision of functions such as site (network) mobility for an entire
   end site [ABH09a].






Atkinson & Bhatti             Experimental                      [Page 4]

RFC 6748                        ILNP ADV                   November 2012


   This document discusses several such optional advanced deployment
   scenarios for ILNP.  These typically use an ILNP-capable Site Border
   Router (SBR).

   Nothing in this document is a requirement for any ILNP implementation
   or any ILNP deployment.

   Readers are strongly advised to first read the ILNP Architecture
   Description [RFC6740], as this document uses the notation and
   terminology described or referenced in that document.

1.1.  Document Roadmap



   This document describes engineering and implementation considerations
   that are common to ILNP for both IPv4 and IPv6.

   The ILNP architecture can have more than one engineering
   instantiation.  For example, one can imagine a "clean-slate"
   engineering design based on the ILNP architecture.  In separate
   documents, we describe two specific engineering instances of ILNP.
   The term "ILNPv6" refers precisely to an instance of ILNP that is
   based upon, and backwards compatible with, IPv6.  The term "ILNPv4"
   refers precisely to an instance of ILNP that is based upon, and
   backwards compatible with, IPv4.

   Many engineering aspects common to both ILNPv4 and ILNPv6 are
   described in [RFC6741].  A full engineering specification for either
   ILNPv6 or ILNPv4 is beyond the scope of this document.

   Readers are referred to other related ILNP documents for details not
   described here:

   a) [RFC6740] is the main architectural description of ILNP, including
      the concept of operations.

   b) [RFC6741] describes engineering and implementation considerations
      that are common to both ILNPv4 and ILNPv6.

   c) [RFC6742] defines additional DNS resource records that support
      ILNP.

   d) [RFC6743] defines a new ICMPv6 Locator Update message used by an
      ILNP node to inform its correspondent nodes of any changes to its
      set of valid Locators.







Atkinson & Bhatti             Experimental                      [Page 5]

RFC 6748                        ILNP ADV                   November 2012


   e) [RFC6744] defines a new IPv6 Nonce Destination Option used by
      ILNPv6 nodes (1) to indicate to ILNP correspondent nodes (by
      inclusion within the initial packets of an ILNP session) that the
      node is operating in the ILNP mode and (2) to prevent off-path
      attacks against ILNP ICMP messages.  This Nonce is used, for
      example, with all ILNP ICMPv6 Locator Update messages that are
      exchanged among ILNP correspondent nodes.

   f) [RFC6745] defines a new ICMPv4 Locator Update message used by an
      ILNP node to inform its correspondent nodes of any changes to its
      set of valid Locators.

   g) [RFC6746] defines a new IPv4 Nonce Option used by ILNPv4 nodes to
      carry a security nonce to prevent off-path attacks against ILNP
      ICMP messages and also defines a new IPv4 Identifier Option used
      by ILNPv4 nodes.

   h) [RFC6747] describes extensions to Address Resolution Protocol
      (ARP) for use with ILNPv4.

1.2.  Terminology



   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.  Localised Numbering



   Today, Network Address Translation (NAT) [RFC3022] is used for a
   number of purposes.  Whilst one of the original intentions of NAT was
   to reduce the rate of use of global IPv4 addresses, through use of
   IPv4 private address space [RFC1918], NAT also offers to site
   administrators a convenient localised address management capability
   combined with a local-scope/private address space, for example,
   [RFC1918] for IPv4.

   For IPv6, NAT would not necessarily be required to reduce the rate of
   IPv6 address depletion, because the availability of addresses is not
   such an issue as for IPv4.  The IETF has standardised Unique Local
   IPv6 Unicast Addresses [RFC4193], which provide local-scope IPv6
   unicast address space that can be used by end sites.  However,
   localised address management, in a manner similar to that provided by









Atkinson & Bhatti             Experimental                      [Page 6]

RFC 6748                        ILNP ADV                   November 2012


   IPv4 NAT and private address space [RFC1918], is still desirable for
   IPv6 [RFC5902], even though there is debate about the efficacy of
   such an approach [RFC4864].

   One of the major concerns that many have had with NAT is the loss of
   end-to-end transport-layer and network-layer session state
   invariance, which is still considered an important architectural
   principle by the IAB [RFC4924].  Nevertheless, the use of localised
   addressing remains in wide use and there is interest in its continued
   use in IPv6, e.g., proposals such as [RFC6296].

   It is possible to have the benefits of NAT-like functions for ILNP
   without losing end-to-end state.  Indeed, such a mechanism -- the use
   of Locator rewriting in ILNP -- forms the basis of many of the
   optional functions described in this document.  In ILNP, we call this
   feature "localised numbering".

   Recall, that a Locator value in ILNP has the same semantics as a
   routing prefix in IP: indeed, in ILNPv4 and ILNPv6 [RFC6741], routing
   prefixes from IPv4 and IPv6, respectively, are used as Locator
   values.

   We note that a deployment using private/local numbering can also
   provide a convenient solution to centralised management of site
   multihoming and network mobility by deploying SBRs in this manner --
   this is described below.

   Please note that with this proposal, localised numbering (e.g., using
   the equivalent of IP NAT on the ILNP Locator bits) would work in
   harmony with multihoming, mobility (for individual hosts and whole
   networks), and IP Security (IPsec), plus the other advanced functions
   described in this document [BA11] [LABH06] [ABH07a] [ABH07b] [ABH08a]
   [ABH08b] [ABH09a] [ABH09b] [RAB09] [RB10] [ABH10] [BAK11].

2.1.  Localised Locators



   For ILNP, the NAT-like function can best be descried by using a
   simple example, based on Figure 2.1.













Atkinson & Bhatti             Experimental                      [Page 7]

RFC 6748                        ILNP ADV                   November 2012


          site                         . . . .      +----+
         network        SBR           .       .-----+ CN |
         . . . .      +------+ L_1   .         .    +----+
        .       .     |      +------.           .
       .         .L_L |      |      .           .
       .         .----+      |      . Internet  .
       .  H      .    |      |      .           .
        .       .     |      |      .           .
         . . . .      +------+       .         .
                                      .       .
                                       . . . .

            CN = Correspondent Node
             H = Host
           L_1 = global Locator value
           L_L = local Locator value
           SBR = Site Border Router

   Figure 2.1: A Simple Localised Numbering Example for ILNP

   In this scenario, the SBR is allocated global locator value L_1 from
   the upstream provider.  However, the SBR advertises internally a
   "local" Locator value L_L.  By "local" we mean that the Locator value
   only has significance within the site network, and any packets that
   have L_L as a source Locator cannot be forwarded beyond the SBR with
   value L_L as the source Locator.  In engineering terms, L_L would,
   for example, in ILNPv6, be an IPv6 prefix based on the assignments
   possible according to IPv6 Unique Local Addresses (ULAs) [RFC4193].

   If we assume that H uses Identifier I_H, then it will use Identifier-
   Locator Vector (I-LV) [I_H, L_L], and that the correspondent node
   (CN) uses IL-V [I_CN, L_CN].  If we consider that H will send a UDP
   packet from its port P_H to CN's port P_CN, then H could send a
   UDP/ILNP packet with the tuple expression:

     <UDP: I_H, I_CN, P_H, P_CN><ILNP: L_L, L_CN>           --- (1a)

   When this packet reaches the SBR, it knows that L_L is a local
   Locator value and so rewrites the source Locator on the egress packet
   to L_1 and forwards that out onto its external-facing interface.  The
   value L_1 is a global prefix, which allows the packet to be routed
   globally:

     <UDP: I_H, I_CN, P_H, P_CN><ILNP: L_1, L_CN>           --- (1b)

   This packet reaches CN using normal routing based on the Locator
   value L_1, as it is a routing prefix.




Atkinson & Bhatti             Experimental                      [Page 8]

RFC 6748                        ILNP ADV                   November 2012


   Note that from expressions (1a) and (1b), the end-to-end state (in
   the UDP tuple) remains unchanged -- end-to-end state invariance is
   honoured, for UDP.  CN would send a UDP packet to H as:

     <UDP: I_CN, I_H, P_CN, P_H><ILNP: L_CN, L_1>           --- (2a)

   and the SBR would rewrite the Locator value on the ingress packet
   before forwarding the packet on its internal interface:

     <UDP: I_CN, I_H, P_CN, P_H><ILNP: L_CN, L_L>           --- (2b)

   Again, this preserves the end-to-end transport-layer session state
   invariance.

   As the Locator values are not used in the transport-layer pseudo-
   header for ILNP [RFC6741], the checksum would not have to be
   rewritten.  That is, the Locator rewriting function is stateless and
   has low overhead.

   (A discussion on the generation of Identifier values for initial use
   is presented in [RFC6741].)

2.2.  Mixed Local/Global Numbering



   It is possible for the SBR to advertise both L_1 and L_L within the
   site, and for hosts within the site to have IL-Vs using both L_1 and
   L_L.  For example, host H may have IL-Vs [I_H, L_1] and [I_H, L_L].
   The configuration and use of such a mechanism can be controlled
   through local policy.

2.3.  Dealing with Internal Subnets with Locator Rewriting



   Where the site network uses subnets, packets will need to be routed
   correctly, internally.  That is, the site network may have several
   internal Locator values, e.g., L_La, L_Lb, and L_Lc.  When an ingress
   packet has I-LV [I_H, L_1], it is expected that the SBR is capable of
   identifying the correct internal network for I_H, and so the correct
   Locator value to rewrite for the ingress packet.  This is not obvious
   as the I value and the L value are not related in any way.

   There are numerous ways the SBR could facilitate the correct lookup
   of the internal Locator value.  This document does not prescribe any
   specific method.  Of course, we do not preclude mappings directly
   from Identifier values to internal Locator values.

   Of course, such a "flat" mapping (between Identifier values and
   Locators) would serve, but maintaining such a mapping would be
   impractical for a large site.  So, we propose the following solution.



Atkinson & Bhatti             Experimental                      [Page 9]

RFC 6748                        ILNP ADV                   November 2012


   Consider that the Locator value, L_x consists of two parts, L_pp and
   L_ss, where L_pp is a network prefix and L_ss is a subnet selector.
   Also, consider that this structure is true for both the local
   identifier, L_L, as well as the global Identifier, L_1.  Then, an SBR
   need only know the mapping from the values of L_ss as visible in L_1
   and the values of L_ss used locally.

   Such a mapping could be mechanical, e.g., the L_ss part of L_L and
   L_1 are the same and it is only the L_pp part that is different.
   Where this is not desirable (e.g., for obfuscation of interior
   topology), an administrator would need to configure a suitable
   mapping policy in the SBR, which could be realised as a simple lookup
   table.  Note that with such a policy, the L_pp for L_L and L_1 do not
   need to be of the same size.

   From a practical perspective, this is possible for both ILNPv6
   [RFC6177] and ILNPv4 [RFC4632].  For ILNPv6, recall that the Locator
   value is encoded to be syntactically similar to an IPv6 address
   prefix, as shown in Figure 2.2, taken from [RFC6741].

   /* IPv6 */
   | 3 |     45 bits         |  16 bits  |     64 bits             |
   +---+---------------------+-----------+-------------------------+
   |001|global routing prefix| subnet ID |  Interface Identifier   |
   +---+---------------------+-----------+-------------------------+
   /* ILNPv6 */
   |             64 bits                 |     64 bits             |
   +---+---------------------+-----------+-------------------------+
   |          Locator (L64)              |  Node Identifier (NID)  |
   +---+---------------------+-----------+-------------------------+
   +<-------- L_pp --------->+<- L_ss -->+

     L_pp = Locator prefix part (assigned IPv6 prefix)
     L_ss = Locator subnet selector (locally managed subnet ID)

   Figure 2.2: IPv6 Address format [RFC3587] as used in ILNPv6, showing
   how subnets can be identified.

   Note that the subnet ID forms part of the Locator value.  Note also
   that [RFC6177] allows the global routing prefix to be more than 45
   bits, and for the subnet ID to be smaller, but still preserving the
   64-bit size of the Locator overall.

   For ILNPv4, the L_pp value overall is an IPv4 routing prefix, which
   is typically less than 32 bits.  However, the ILNPv4 Locator value is
   carried in the 32-bit IP Address space, so the bits not used for the





Atkinson & Bhatti             Experimental                     [Page 10]

RFC 6748                        ILNP ADV                   November 2012


   routing prefix could be used for L_ss, e.g., for a /24 IPv4 prefix,
   the situation would be as shown in Figure 2.3, and L_ss could use any
   of the remaining 8-bits as required.

              24 bits           8 bits
     +------------------------+----------+
     |         Locator (L32)             |
     +------------------------+----------+
     +<------- L_pp --------->+<- L_ss ->+

     L_pp = Locator prefix (assigned IPv4 prefix)
     L_ss = Locator subnet selector (locally managed subnet ID)

   Figure 2.3: IPv4 address format for /24 IPv4 prefix, as used in
   ILNPv4, showing how subnets can be identified.

   As an example, for the case where the interior topology is not
   obfuscated, an interior "engineering" node might have an LP record
   pointing to eng.example.com and eng.example.com might have L32/L64
   records for a specific subnet inside the site.  Meanwhile, an
   interior "operations" node might have an LP record pointing at
   "ops.example.com" that might have different L32/L64 records for that
   specific subnet within the site.  That is, eng.example.com might have
   Locator value L_pp_1:L_ss_1 and ops.example.com might have Locator
   value L_pp_1:L_ss_2.  However, just as for IPv6 or IPv4 routing
   today, the routing for the site would only need to use L_pp_1, which
   is a routing prefix in either IPv6 (for ILNPv6) or IPv4 (for ILNPv4).

2.4.  Localised Name Resolution with DNS



   To support private numbering with IPv4 and IPv6 today, some sites use
   a split-horizon DNS service for the site [appDNS].

   If a site using localised numbering chooses to deploy a split-horizon
   DNS server, then the DNS server would return the global-scope
   Locator(s) (L_1 in our example above) of the SBR to DNS clients
   outside the site, and would advertise the local-scope Locator(s) (L_L
   in our example above) specific to that internal node to DNS clients
   inside the site.  Such deployments of split-horizon DNS servers are
   not unusual in the IPv4 Internet today.  If an internal node (e.g.,
   portable computer) moves outside the site, it would follow the normal
   ILNP methods to update its authoritative DNS server with its current
   Locator set.  In this deployment model, the authoritative DNS server
   for that mobile device will be either the split-horizon DNS server
   itself or the master DNS server providing data to the split-horizon
   DNS server.





Atkinson & Bhatti             Experimental                     [Page 11]

RFC 6748                        ILNP ADV                   November 2012


   If a site using localised numbering chooses not to deploy a split-
   horizon DNS server, then each internal node would advertise the
   global-scope Locator(s) of the site border routers in its respective
   DNS entries.  To deliver packets from one internal node to another
   internal node, the site would choose to use either Layer 2 bridging
   (e.g., IEEE Spanning Tree or IEEE Rapid Spanning Tree [IEEE04], or a
   link-state Layer 2 algorithm such as the IETF TRILL group or IEEE
   802.1 are developing), or the interior routers would forward packets
   up to the nearest site border router, which in turn would then
   rewrite the Locators to appropriate local-scope values, and forward
   the packet towards the interior destination node.

   Alternately, for sites using localised numbering but not deploying a
   split-horizon DNS server, the DNS server could return all global-
   scope and local-scope Locators to all queriers, and assume that nodes
   would use normal, local address/route selection criteria to choose
   the best Locator to use to reach a given remote node ([RFC3484] for
   older IPv6 nodes, [RFC6724] for newer IPv6 nodes).  Hosts within the
   same site as the correspondent node would only have a ULA configured;
   hence, they would select the ULA destination Locator for the
   correspondent (L_L in our example).  Hosts outside the site would not
   have the same ULA configured (L_CN for the CN in our example).

   However, ILNP allows use of Locator Preference values [RFC6742]
   [RFC6743].  These values would indicate explicitly the relative
   preference value given to Locator values and so result in the
   selection of the appropriate Locator (and therefore interface) to use
   for the transmission of an outgoing packet with respect to the value
   to be inserted into the IPv6 Source Address field (see Section 3 of
   [RFC6741]).  A similar argument, with respect to use of Locator
   preference values, applies to the value to be inserted into the IPv6
   Destination Address field.  Certainly, by using appropriate
   Preference values for a host with multiple Locator values, it would
   be possible to emulate some level of resemblance to the address
   selection rules in [RFC3484] and [RFC6724], and this could be
   controlled via DNS entries for ILNP nodes, for example.

   Indeed, with appropriate use of localised or site-wide policy, and
   appropriate mechanisms in the devices (e.g. in end hosts operating
   systems or in Site Border Routers), Preference values for Locator
   values within the DNS could be used for allowing options for multi-
   homed transport sessions and/or site-controlled traffic engineering
   [ABH09a].  However, the details for this are left for further study,
   and overall, the rules defined in [RFC3484] and [RFC6724] cannot be
   applied directly to ILNPv6 nodes.






Atkinson & Bhatti             Experimental                     [Page 12]

RFC 6748                        ILNP ADV                   November 2012


   Note that for split-horizon operation, there needs to be a DNS
   management policy for mobile hosts, as when such hosts are away from
   their "home" network, they will need to update DNS entries so that
   the global-scope Locator(s) only is (are) used, and these are
   consistent with the current topological position of the mobile host.
   Such updates would need to be done using Secure Dynamic DNS Update.

   For an ILNP mobile network using LP records, there are likely to
   separate LP records for internal and external use.

2.5.  Use of mDNS



   Multicast DNS (mDNS) [mDNS11] is popularly used in many end-system
   OSs today, especially desktop OSs (such as Windows, Mac OS X and
   Linux).  It is used for localised name resolution using names with a
   ".local" suffix, for both IPv4 and IPv6.  This protocol would need to
   be modified so that when an ILNP-capable node advertises its ".local"
   name, another ILNP-capable node would be able to see that it is an
   ILNP-capable, but other, non-ILNP nodes would not be perturbed in
   operation.  The details of a mechanism for using mDNS to enable such
   a feature are not defined here.

2.6.  Site Network Name in DNS



   In this scenario, if H expects incoming ILNP session requests, for
   example, then remote nodes normally will need to look up appropriate
   Identifier and Locator information in the DNS.  Just as for IP, and
   as already described in [RFC6740], a Fully Qualified Domain Name
   (FQDN) lookup for H should resolve to the correct NID and L32/L64
   records.  If there are many hosts like H that need to keep DNS
   records (for any reason, including to allow incoming ILNP session
   requests), then, potentially, there are many such DNS resource
   records.

   As an optimisation, the network as a whole may be configured with one
   or more L32 and L64 records (to store the value L_1 from our example)
   that are resolved from an FQDN.  At the same time, individual hosts
   now have an FQDN that returns one or more LP record entries [RFC6742]
   as well as NID records.  The LP record points to the L32 or L64
   records for the site.  A multihomed site normally will have at least
   one L32 or L64 record for each distinct uplink (i.e., link from a
   Site Border Router towards the global Internet), because ILNP uses
   provider-aggregatable addressing.

   More than one L32 or L64 will be required if multiple Locator values
   are in use.  For example, if an ILNPv6 site has multiple links for
   multihoming, it will use one L64 record for each Locator value it is
   using on each link.



Atkinson & Bhatti             Experimental                     [Page 13]

RFC 6748                        ILNP ADV                   November 2012


2.7.  Site Interior Topology Obfuscation



   In some situations, it can be desirable to obfuscate the details of
   the interior topology of an end site.  Alternately, in some
   situations, local site policy requires that local-scope routing
   prefixes be used within the local site.  ILNP can provide these
   capabilities through the ILNP local addressing capability described
   here, under the control of the SBR.

   As described in Section 2.3 above, locator rewriting can be used to
   hide the internal structure of the network with respect to the
   subnetting arrangement of the site network.  Specifically, the
   procedure described in Section 2.3 would be followed, with the
   following additional modification of the use of Locator values:

   (1) Only the aggregated Locator value, i.e., L_pp, is advertised
       outside the site (e.g., in an L32 or L64 record), and L_ss is
       zeroed in that advertisement.

   (2) The SBR needs to maintain a mapping table to restore the interior
       topology information for received packets, for example, by using
       a mapping table from I values to either L_ss values or internal
       Locator values.

   (3) The SBR needs to zero the L_ss values for all Source Locators of
       egress packets, as well as perform a Locator rewriting that
       affects the L_pp bits of the Locator value.

   Of course, this only obscures the interior topology of the site, not
   the exterior connectivity of the site.  In order for the site to be
   reachable from the global Internet, the site's DNS entries need to
   advertise Locator values for the site to the global Internet (e.g.,
   in L32, L64 records).

2.8.  Other SBR Considerations



   For backwards compatibility, for ILNP, the ICMP checksum is always
   calculated identically as for IPv6 or IPv4.  For ILNPv6, this means
   that the SBR need not be aware if ILNPv6 is operating as described in
   [RFC6740] and [RFC6741].  For ILNPv4, again, the SBR need not be
   aware of the operation if ILNPv4 is operating as it will not need to
   inspect the extension header carrying the I value.

   In order to support communication between two internal nodes that
   happen to be using global-scope addresses (for whatever reason), the
   SBR MUST support the "hair pinning" behaviour commonly used in
   existing NAT/NAPT devices.  (This behaviour is described in Section 6
   of RFC 4787 [RFC4787].)



Atkinson & Bhatti             Experimental                     [Page 14]

RFC 6748                        ILNP ADV                   November 2012


   In the near-term, a more common deployment scenario will be to deploy
   ILNP incrementally, with some ordinary classic IP traffic still
   existing.  In this case, the SBR should maintain flow state that
   contains a flag for each flow indicating whether or not that flow is
   using ILNP.  If that flag indicated ILNP were enabled for a given
   flow, and ILNP local numbering were also enabled, then the SBR would
   know that it should perform the simpler ILNP Locator rewriting
   mapping.  If that flag indicated ILNP were not enabled for a given
   flow and IP NAT or IP NAPT were also enabled, then the SBR would know
   that it should perform the more complex NAT/NAPT translation (e.g.,
   including TCP or UDP checksum recalculation).

      NOTE: Existing commercial security-aware routers (e.g., Juniper
      SRX routers) already can maintain flow state for millions of
      concurrent IP flows.  This feature would add one flag to each
      flow's state, so this approach is believed scalable today using
      existing commercial technology.

   Those applications that do not use IP Address values in application
   state or configuration data are considered to be "well behaved".  For
   well-behaved applications, no further enhancements are required.
   Where application-layer protocols are not well behaved, for example,
   the File Transfer Protocol (FTP), then the SBR might need to perform
   additional stateful processing -- just as NAT and NAPT equipment
   needs to do today for FTP.  See the description in Section 7.6 of
   [RFC6741].

   When the SBR rewrites a Locator in an ILNP packet, that obscures
   information about how well a particular path is working between the
   sender and the receiver of that ILNP packet.  So, the SBR that
   rewrites Locator values needs to include mechanisms to ensure that
   any packet with a new Destination Locator will travel along a valid
   path to the intended destination node.  For ILNPv4, the path liveness
   will be no worse than IPv4, and mechanisms already in use for IPv4
   can be reused.  For ILNPv6, the path liveness will be no worse than
   for IPv6, and mechanisms already in use for IPv6 can be reused.

   In the future, the Border Router Discovery Protocol (BRDP) also might
   be used in some deployments to indicate which routing prefixes are
   currently valid and which site border routers currently have a
   working uplink [BRDP11].










Atkinson & Bhatti             Experimental                     [Page 15]

RFC 6748                        ILNP ADV                   November 2012


3.  An Alternative for Site Multihoming



   The ILNP Architectural Description [RFC6740] describes the basic
   approach to enabling Site Multihoming (S-MH) with ILNP.  However, as
   an option, it is possible to leave the control of S-MH to an ILNP-
   enabled SBR.  This alternative is based on the use of the Localised
   Numbering function described in Section 2 of this document.

3.1.  Site Multihoming (S-MH) Connectivity Using an SBR



   The approach to Site Multihoming (S-MH) using an SBR is best
   illustrated through an example, as shown in Figure 3.1.

          site                         . . . .      +----+
         network         SBR          .       .-----+ CN |
         . . . .      +------+ L_1   .         .    +----+
        .       .     |  sbr1+------.           .
       .         .L_L |      |      .           .
       .         .----+      |      . Internet  .
       .  H      .    |      |      .           .
        .       .     |  sbr2+------.           .
         . . . .      +------+ L_2   .         .
                                     .       .
                                      . . . .

             CN = Correspondent Node
              H = Host
            L_1 = global Locator value 1
            L_2 = global Locator value 2
            L_L = local Locator value
            SBR = Site Border Router
           sbrN = interface N on SBR

    Figure 3.1: Alternative Site Multihoming Example with an SBR

   The situation here is similar to the localised numbering example,
   except that the SBR now has two external links, with using Locator
   value L_1 and another using Locator value L_2.  These could, e.g.,
   for ILNPv6, be separate, Provider Aggregated (PA) IPv6 prefixes from
   two different ISPs.  H has IL-V [I_H, L_L], and will forward a packet
   to CN as given in expression (1a).  However, when the packet reaches
   the SBR, local policy will decide whether the packet is forwarded on
   the link sbr1 using L_1 or on sbr2 using L_2.  Of course, the correct
   Locator value will be rewritten into the egress packet in place of
   L_L.






Atkinson & Bhatti             Experimental                     [Page 16]

RFC 6748                        ILNP ADV                   November 2012


   If only local numbering is being used, then the SBR need never
   advertise any global Locator values.  However, it could do, as
   described in Section 2.2.

3.2.  Dealing with Link/Connectivity Changes



   One of the key uses for multihoming is providing resilience to link
   failure.  If either link breaks, then the SBR can manage the change
   in connectivity locally.  For example, assume SBR has been configured
   to use sbr1 for all traffic, and sbr2 only as backup link.  So, SBR
   directs packets from H to communicate with CN using sbr1, and CN will
   receive packets as in expression (1b) and respond with packets as in
   expression (2a).

   However, if sbr1 goes down then SBR will move the communication to
   interface sbr2.  As H is not aware of the actions of the SBR, the SBR
   must maintain some state about IL-V "pairs" in order to hand off the
   connectivity from sbr1 to sbr2.  So, when moving the communication to
   sbr2, the SBR would firstly send a Locator Update (LU) message
   [RFC6745] [RFC6743], to CN informing it that L_2 is now the valid
   Locator for the communication.  This operation would not be visible
   to H, although there might be some disruption to transmission, e.g.,
   packets being sent from CN to H that are in flight when sbr1 goes
   down may be lost.  The SBR might also need to update DNS entries (see
   Section 3.3).  Since ILNP requires that all Locator Update messages
   be authenticated by the ILNP Nonce, the SBR will need to include the
   appropriate Nonce values as part of its cache of information about
   ILNP sessions traversing the SBR.  (NOTE: Since commercial security
   gateways available as of this writing reportedly can handle full
   stateful packet inspection for millions of flows at multi-gigabit
   speeds, it should be practical for such devices to cache the ILNP
   flow information, including Nonce values.)

   This approach has some efficiency gains over the approach for
   multihoming described in [RFC6740], where each hosts manages its own
   connectivity.

   If sbr1 was to be reinstated, now with Locator value L_3, then local
   policy would determine if the communication should be moved back to
   sbr1, with appropriate additional actions, such as transmission of LU
   messages with the new Locator values and also the updates to DNS.

   Note that in such movement of an ILNP session across interfaces at
   the SBR, only Locator values in ILNP packets are changed.  As already
   noted in [RFC6740], end-to-end transport-layer session state
   invariance is maintained.





Atkinson & Bhatti             Experimental                     [Page 17]

RFC 6748                        ILNP ADV                   November 2012


3.3.  SBR Updates to DNS



   When the SBR manages connectivity as described above, the internal
   hosts, such as H, are not necessarily aware of any connectivity
   changes.  Indeed, there is certainly no requirement for them to be
   aware.  So, if H was a server expecting incoming connections, the SBR
   must update the relevant DNS entries when the site connectivity
   changes.

   There are two possibilities: each host could have its own L32 or L64
   records; or the site might use a combination of LP and L32/L64
   records (see Section 2.4).  Either way, the SBR would need to update
   the relevant DNS entries.  For our example, with ILNPv6 and LP
   records in use, the SBR would need to manage two L64 records (one for
   each uplink) that would resolve from a FQDN, for example,
   site.example.com.  Meanwhile, individual hosts, such as H, have an
   FQDN that resolves to an NID value and an LP record that would
   contain the value site.example.com, which then would be used to look
   up the two L64 records.

   If the SBR is multihomed, as in Figure 3.1, then it will have (at
   least) two Locator values, one for each link, and local policy will
   need to be used to determine how preference values are applied in the
   relevant L32 and L64 records.

3.4.  DNS TTL Values for L32 and L64 Records



   Imagine that in the scenario described above, there was a link
   failure that resulted in sbr1 going down and sbr2 was used.  Existing
   ILNP sessions in progress would move to sbr2 as described above.
   However, new incoming ILNP sessions to the site would need to know to
   use L_2 and not L_1.  L_1 and L_2 would be stored in DNS records
   (e.g., L32 for ILNPv4 or L64 for ILNPv6).  If a remote host has
   already resolved from DNS that L_1 is the correct Locator for sending
   packets to the site, then that host might be holding stale
   information.

   DNS allows values returned to be aged using Time-To-Live (TTL), which
   is specified in the time unit of seconds.  So that remote nodes do
   not hold on to stale values from DNS, the L64 records for our site
   should have low TTL values.  An appropriate value must be considered
   carefully.  For example, let us assume that the site administrator
   knows that when sbr1 fails, it takes 20 seconds to failover to sbr2.
   Then, 20 s would seem to be an appropriate time to use for the TTL
   value of an L64 for the site: if a remote node had just resolved the
   value L_1 for the site, and the link to sbr1 went down, that remote
   node would not hold the stale value of L_1 for any longer than it
   takes the site to failover to sbr2 and use L_2.



Atkinson & Bhatti             Experimental                     [Page 18]

RFC 6748                        ILNP ADV                   November 2012


   Our studies for a university school site network show that low TTL
   values, as low as zero, are feasible for operational use [BA11].

   NOTE: From 01 November 2010, the site network of the School of
         Computer Science, University of St Andrews, UK, has been
         running operational DNS with DNS A records that have TTL of
         zero.  At the time of writing of this document (November 2012),
         a zero DNS TTL was still in use at the school.

3.5.  Multiple SBRs



   For site multihoming, with multiple SBRs, a situation may be as
   follows (see also Section 5.3.1 in [RFC6740]).

         site                          . . . .
        network                       .       .
        . . . .      +-------+ L_1   .         .
       .       .     |       +------.           .
      .         .    |       |      .           .
     .           .---+ SBR_A |      .           .
     .           .   |       |      .           .
     .           .   |       |      .           .
     .           .   +-------+      .           .
     .           .       ^          .           .
     .           .       | CP       . Internet  .
     .           .       v          .           .
     .           .   +-------+ L_2  .           .
     .           .   |       +------.           .
     .           .   |       |      .           .
     .           .---+ SBR_B |      .           .
      .         .    |       |      .           .
       .       .     |       |      .           .
        . . . .      +-------+       .         .
                                      .       .
                                       . . . .

         CP     = coordination protocol
         L_1    = global Locator value 1
         L_2    = global Locator value 2
         SBR_A  = Site Border Router A
         SBR_B  = Site Border Router P

   Figure 3.2: A Dual-Router Multihoming Scenario for ILNP

   The use of two physical routers provides an extra level of resilience
   compared to the scenario of Figure 3.1.  The coordination protocol
   (CP) between the two routers keeps their actions in synchronisation
   according to whatever management policy is in place for the site



Atkinson & Bhatti             Experimental                     [Page 19]

RFC 6748                        ILNP ADV                   November 2012


   network.  Such functions are available today in some commercial
   network security products.  Note that, logically, there is little
   difference between Figures 5.1 and 3.2, but with two distinct routers
   in Figure 3.2, the interaction using CP is required.  Of course, it
   is also possible to have multiple interfaces in each router and more
   than two routers.

4.  An Alternative for Site (Network) Mobility



   The ILNP Architectural Description [RFC6740] describes the basic
   approach to enabling site (network) mobility with ILNP.  However, as
   an option, it is possible to leave the control of site mobility to an
   ILNP-enabled SBR by exploiting the alternative site multihoming
   feature described in Section 3 of this document.

   Again, as described in [RFC6740], we exploit the duality between
   mobility and multihoming for ILNP.

4.1.  Site (Network) Mobility



   Let us consider the mobile network in Figure 4.2, which is taken from
   [RFC6740].

          site                        ISP_1
         network        SBR           . . .
         . . . .      +------+ L_1   .     .
        .       . L_L |   ra1+------.       .
       .         .----+      |      .       .
        .  H    .     |   ra2+--    .       .
         . . . .      +------+       .     .
                                      . . .

       Figure 4.1a: ILNP Mobile Network before Handover

          site                        ISP_1
         network        SBR           . . .
         . . . .      +------+ L_1   .     .
        .       . L_L |   ra1+------. . . . .
       .         .----+      |      .       .
        .  H    .     |   ra2+------.       .
         . . . .      +------+ L_2  . . . . .
                                     .     .
                                      . . .
                                      ISP_2

       Figure 4.1b: ILNP Mobile Network during Handover





Atkinson & Bhatti             Experimental                     [Page 20]

RFC 6748                        ILNP ADV                   November 2012


          site                        ISP_2
         network        SBR           . . .
         . . . .      +------+       .     .
        .       . L_L |   ra1+--    .       .
       .         .----+      |      .       .
        .  H    .     |   ra2+------.       .
         . . . .      +------+ L_2   .     .
                                      . . .

       Figure 4.1c: ILNP Mobile Network after Handover

            H = host
          L_1 = global Locator value 1
          L_2 = global Locator value 2
          L_L = local Locator value
          raN = radio interface N
          SBR = Site Border Router

     Figure 4.1: An Alternative Mobile Network Scenario with an SBR

   We assume that the site (network) is mobile, and the SBR has two
   radio interfaces, ra1 and ra2.  In the figure, ISP_1 and ISP_2 are
   separate, radio-based service providers, accessible via interfaces
   ra1 and ra2.

   While the SBR makes the transition from using a single link (Figure
   4.1a) to the handover overlap on both links (Figure 4.1b), to only
   using a single link again (Figure 4.1c), the host H continues to use
   only Locator value L_L, as already described for Site Multihoming
   (S-MH).  During this time the actions taken by the SBR are the same
   as already described in [RFC6740], except that the SBR:

   a) also performs that ILNP localised numbering function described in
      Section 2.

   b) does not need to advertise L_1 and L_2 internally if only local
      numbering is being used.

   As for the case of S-MH above, H need not be aware of the change in
   connectivity for the SBR if it is only using local numbering, and the
   SBR would send LU messages for H (for any correspondent nodes, not
   shown in Figure 4.1), and would update DNS entries as required.

   The difference to the S-MH scenario described earlier in this
   document is that in the situation of Figure 4.1b, the SBR can opt to
   use soft handover has previously described in [RFC6740].





Atkinson & Bhatti             Experimental                     [Page 21]

RFC 6748                        ILNP ADV                   November 2012


   Again, there is an efficiency gain compared to the situation
   described in [RFC6740]: the SBR provides a convenient point at which
   to centrally manage the movement of the site as a whole.  Note that
   in Figure 4.1b, the site is multihomed.

   As for S-MH, L_1 and L_2 could be advertised internally, as a local
   policy decision, for those hosts that require direct control of their
   connectivity.

   Note that for handover, immediate handover will have a similar
   behaviour to a link outage as described for S-MH.  However, as ILNP
   allows soft-handover, during the handover period, this should help to
   reduce (perhaps even remove) packet loss.

4.2.  SBR Updates to DNS



   As for S-MH, a similar discussion to Section 3.3 applies for mobile
   networks with respect to the updates to DNS.  As a mobile network is
   likely to have more frequent changes to its connectivity than a
   multihomed network would due to connectivity changes, the use of LP
   DNS records is likely to be particularly advantageous here.

4.3.  DNS TTL Values for L32 and L64 Records



   As for S-MH, a similar discussion to Section 3.4 applies for mobile
   networks with respect to the TTL of L32 and/or L64 records that are
   used for the name of the mobile network.  In the case of the mobile
   network, it makes sense for the TTL to be aligned to the time for
   handover.

5.  Traffic Engineering Options



   The use of Locator rewriting provides some simple yet useful options
   for traffic engineering (TE) controlled from the edge-site via the
   SBR, requiring no cooperation from the service provider other than
   the provision of basic connectivity services, e.g., physical
   connectivity, allocation of IP Address prefixes and packet
   forwarding.  This does not preclude other TE options that are already
   in use, such as use of MPLS, but we choose to highlight here the
   specific options available and controllable solely through the use of
   ILNP.

   When a site network is multihomed, we have seen that the use of the
   Locator rewriting function permits the SBR to have packet-by-packet
   control when forwarding on external links.  Various configuration and
   policies could be applied at the SBR in order to control the egress
   and ingress traffic to the site network.




Atkinson & Bhatti             Experimental                     [Page 22]

RFC 6748                        ILNP ADV                   November 2012


5.1.  Load Balancing



   Let us consider Figure 5.1, and assume ILNP local numbering is in
   use; that H1, H2, and H3 use, respectively, Identifier values, I_1,
   I_2 and I_3; and all of them use Locator value L_L.

           site                         . . . .
          network         SBR          .       .
          . . . .      +------+ L_1   .         .
         .       .     |  sbr1+------.           .
        .     H2  .L_L |      |      .           .
        . H3      .----+      |      . Internet  .
        .         .    |      |      .           .
         .  H1   .     |  sbr2+------.           .
          . . . .      +------+ L_2   .         .
                                       .       .
                                        . . . .

            HN = host N
           L_1 = global Locator value 1
           L_2 = global Locator value 2
           L_L = local Locator value
           SBR = Site Border Router
          sbrN = interface N on sbr

      Figure 5.1: A Site Multihoming Scenario for Traffic Control

   The SBR could be configured, subject to local policy, to try to
   control load across the external links.  For example, it could be
   configured initially with the following mappings:

     srcI=I_1, sbr1                                        --- (3a)
     srcI=I_2, sbr2                                        --- (3b)
     srcI=I_3, sbr1                                        --- (3c)

   These mappings direct packets matching course Identifier values to
   particular outgoing interfaces.  As load changes, these mappings
   could be changed.  For example, expression (3c) could be changed to:

     srcI=I_3, sbr2                                        --- (4)

   and the SBR would need to send LU message to the correspondents of H3
   (sbr to uses L_2 while sbr1 uses L_1).  The egress connectivity is
   totally within control of the SBR under administrative policy, as
   already seen in the descriptions of multihoming and mobility in this
   document.





Atkinson & Bhatti             Experimental                     [Page 23]

RFC 6748                        ILNP ADV                   November 2012


   Of course, more complex policies are possible, based on:

    - whether ILNP sessions are incoming or outgoing
    - time of day
    - internal subnets

   and any number of criteria already in use for control of traffic.

   In expressions (3a,b,c) above, source I values are used.  However:

    - destination I values could be used
    - source or destination L values could be used
    - mappings could be to L values, not to specific interfaces

   and, again, any number of criteria could be used to manipulate the
   packet path, based on filtering of values in header fields and local
   policy.

   With ILNP, hosts do not need to be aware of the operation of the SBR
   in this manner.

   Note, again, that in this scenario, there is nothing to prevent SBR
   from also advertising L_1 and L_2 into the site network.  If
   required, administrative controls could be used to enable selective
   hosts in the site network to use L_1 and L_2 directly as described in
   [RFC6740].

5.2.  Control of Egress Traffic Paths



   Extending the scenario for load-balancing described above, it is also
   be possible for the ILNP-capable SBR to direct traffic along specific
   network paths based on the use of different L values, i.e., by using
   multiple prefixes assigned from upstream providers.

   Of course, as previously discussed, these prefixes can be Provider
   Aggregated (PA) and need not be Provider Independent (PI).

   Let us consider Figure 5.2 and assume ILNP local numbering is in use;
   that H1, H2 and H3 use, respectively, Identifier values, I_1, I_2,
   and I_3; and all of them use Locator value L_L.  Let us also assume
   that the node CN uses IL-V [I_CN, L_CN].










Atkinson & Bhatti             Experimental                     [Page 24]

RFC 6748                        ILNP ADV                   November 2012


           site                           . . . .      +----+
          network         SBR            .       .-----+ CN |
          . . . .      +------+ L1,L2   .         .    +----+
         .       .     |  sbr1+--------.           .
        .     H2  .L_L |      |        .           .
        . H3      .----+  sbr2+--------. Internet  .
        .         .    |      | L3,L4  .           .
        .         .    |      |        .           .
         .  H1   .     |  sbr3+--------.           .
          . . . .      +------+ L5,L6   .         .
                                         .       .
                                          . . . .

            CN = correspondent node
            HN = host N
            LN = global Locator value N
           L_L = local Locator value
           SBR = Site Border Router
          sbrN = interface N on sbr

      Figure 5.2: A Site Multihoming Scenario for Traffic Control

   Here, many configurations are possible.  For example, for egress
   traffic:

     srcI=I_2, L2                                          --- (5a)
     srcI=I_3, L3                                          --- (5b)
     dstI=I_CN, L6                                         --- (5c)
     srcI=I_1 dstI=I_CN, L1                                --- (5d)

   Expression (5a) maps all egress packets from H2 to have their source
   Locator value rewritten to L2 (and implicitly to use interface sbr1).
   Expression (5b) maps all egress packets from H3 to have their source
   Locator value rewritten to L3 (and implicitly to use interface sbr2).
   Expression (5c) directs any traffic to CN to use Locator value L6 as
   the source Locator (and implicitly to use interface sbr3), and may
   override (5a) and (5b), subject to local policy, when packets to CN
   are from H2 or H3.

   Meanwhile, in expression (5d), we see a further, more specific rule,
   in that packets from H1 destined to CN should use Locator value L1
   (and implicitly to use interface sbr1).

   Note the implicit bindings to interfaces in expressions (5a,b,c,d),
   compared to the explicit bindings in expressions (3a,b,c).  ILNP only
   requires that the Locator values are correctly rewritten and packets
   forwarded in conformance with the routing already configured for the
   Locator values.



Atkinson & Bhatti             Experimental                     [Page 25]

RFC 6748                        ILNP ADV                   November 2012


   Of course, these rules can be changed dynamically at the SBR, and the
   SBR will migrate ILNP sessions across Locator values, as already
   described above for mobility.

6.  ILNP in Datacentres



   As ILNP has first class support for mobility and multihoming, and
   supports flexible options for localised addressing, there is great
   potential for it to be used in datacentre scenarios.  Further details
   of possibilities are in [BA12], with a summary presented here.

   There are several scenarios that could be beneficial to datacentres,
   in order to provide functions such as load balancing, resilience and
   fault tolerance, and resource management:

   - Same datacentre, internal Virtual Machine (VM) mobility: This could
     be beneficial in load balancing, dynamically, where load changes
     are taking place.  The remote user does not see the VM has moved.

   - Different datacentres, transparent mobility: This is where the
     datacentre resources may be geographically distributed, but the
     geographical movement is transparent to the remote user.

   - Different datacentres, mobility is visible: This is where the
     datacentre resources may be geographically distributed, but the
     geographical movement is visible to the remote user.

   These are three situations that may be supported by ILNP, but they
   are not the only ones: we provide these here as examples, and they
   are not intended to be prescriptive.  The intention is only to show
   the flexibility that is possible through the use of ILNP.

   This section describes some Virtual Machine (VM) mobility
   capabilities that are possible with ILNP.  Depending on the internal
   details and virtualisation model provided by a VM platform, it might
   be sufficient for the guest operating system to support ILNP.  In
   some cases, again depending on the internal details and
   virtualisation model provided by a VM platform, the VM platform
   itself also might need to include support for ILNP.

   Details of how a particular VM platform works, and which
   virtualisation model(s) a VM platform supports, are beyond the scope
   of this document.  Internal implementation details of VM platform
   support for ILNP are also beyond the scope of this document, just as
   internal implementation details for any other networked system
   supporting ILNP are beyond the scope of this document.





Atkinson & Bhatti             Experimental                     [Page 26]

RFC 6748                        ILNP ADV                   November 2012


6.1.  Virtual Image Mobility within a Single Datacentre



   Let us consider first the scenario of Figure 6.1, noting its
   similarity to Figure 2.1 for use of localised numbering.

          site                         . . . .      +----+
         network        SBR           .       .-----+ CN |
         . . . .      +------+ L_1   .         .    +----+
        .       .     |      +------.           .
       .    H2   .L_L |      |      .           .
       .         .----+      |      . Internet  .
       .  V*H1   .    |      |      .           .
        .       .     |      |      .           .
         . . . .      +------+       .         .
                                      .       .
                                       . . . .

            CN = Correspondent Node
             V = Virtual machine image
            Hx = Host x
           L_1 = global Locator value
           L_L = local Locator value
          SBR = Site Border Router

     Figure 6.1: A Simple Virtual Image Mobility Example for ILNP

   L_L is a Locator value used for the ILNP hosts H1 and H2.  Here, the
   "V*H1" signifies that the virtual machine image V is currently
   resident on H1.  Let us assume that V has Identifier I_V.  Note that
   as H1 and H2 have the same Locator value (L_1), as far as CN is
   concerned, it does not matter if V is resident on H1 or H2, all
   transport packets between V and CN will have the same signature as
   far as CN is concerned, e.g., for a UDP flow (in analogy to (1a)):

     <UDP: I_V, I_CN, P_V, P_CN><ILNP: L_1, L_CN>           --- (6a)

   Now, if V was to migrate to H2, the migration would be an issue
   purely local to the site network, and the end-to-end integrity of the
   transport flow would be maintained.

   Of course, there are practical operating systems issues in enabling
   such a migration locally, but products exist today that could be
   modified and made ILNP-aware in order to enable such VM image
   mobility.

   Note that for convenience, above, we have used localised numbering
   for ILNP, but if local Locator values were not used and the whole
   site simply used L_1, the principle would be the same.



Atkinson & Bhatti             Experimental                     [Page 27]

RFC 6748                        ILNP ADV                   November 2012


6.2.  Virtual Image Mobility between Datacentres - Invisible



   Let us now consider an extended version of the scenario above in Fig.
   6.2, where we see that there is a second site network, which is
   geographically distant to the first site network, and the two site
   networks are interconnected via their respective SBRs.

          site                         . . . .      +----+
         network 1      SBR1          .       .-----+ CN |
         . . . .      +------+ L_1   .         .    +----+
        .       .     |      +------.           .
       .         .L_L1|      |      .           .
       .         .----+      |      . Internet  .
       .  V*H1   .    |      |      .           .
        .       .     |      |      .           .
         . . . .      +---+--+      .           .
                          :         .           .
                          :         .           .
         . . . .      +---+--+ L_2  .           .
        .       .     |      +------.           .
       .    H2   .L_L2|      |      .           .
       .         .----+      |      .           .
       .         .    |      |      .           .
        .       .     |      |      .           .
         . . . .      +------+       .         .
          site          SBR2          .       .
         network 2                     . . . .

             : = logical inter-router link and coordination
            CN = Correspondent Node
             V = Virtual machine image
            Hx = Host x
           L_y = global Locator value y
          L_Lz = local Locator value z
          SBR = Site Border Router

     Figure 6.2: A Simple Localised Numbering Example for ILNP

   Note that the logical inter-router link between SBR1 and SBR2 could
   be realised physically in many different ways that are available
   today and are not ILNP-specific, e.g., leased line, secure IP-layer
   or Layer 2 tunnel, etc.  We assume that this link also allows
   coordination between the two SBRs.  For now, we ignore external link
   L_2 on SBR2, and assume that the remote node, CN, is in communication
   with V through SBR1.






Atkinson & Bhatti             Experimental                     [Page 28]

RFC 6748                        ILNP ADV                   November 2012


   When in initial communication, the packets have the signature is
   given in expression (6a).  When V moves to H2, it now uses Locator
   value L_L2, but all communication between V and CN is still routed
   via SBR1.  So, the remote CN still sees that same packet signature as
   given in expression (6a).  L_L1 and L_L2 are, effectively, two
   internal (private) subnetworks, and are not visible to CN.

   However, SBR2 and SBR1 must coordinate so that any further
   communication to V via SBR1 is routed across the inter-router link.
   Again, there are commercial products today that could be adapted to
   manage such shared state.

6.3.  Virtual Image Mobility between Datacentres - Visible



   Clearly, in the scenario of the section above, once V has moved to
   site network 2, it may be beneficial, for a number of reasons, for
   communication to V to be routed via SBR2 rather than SBR1.

   When V moves from site network 1 to site network 2, this visibility
   of mobility could be by V sending ILNP Locator Update messages to the
   CN during the mobility process.  Also, V would update any relevant
   ILNP DNS records, such as L64 records, for new ILNP session requests
   to be routed via SBR2.

   Indeed, let us now consider again Figure 6.2, and assume now that
   Local locators L_L1 and L_L2 are not in use on either site network,
   and each site networks uses its own global Locator value, L_1 and
   L_2, respectively, internally.  In that case, the packet flow
   signature for V when it is in site network 1 as viewed from CN is,
   again as given in expression (6a).  However, when V moves to site
   network 2, it would simply use L_2 as its new Locator, send Locator
   Update messages to CN as would a normal mobile node for ILNP, and
   complete its migration to H2.  Then, CN would see the packet
   signatures as in expression (6b).

     <UDP: I_V, I_CN, P_V, P_CN><ILNP: L_2, L_CN>           --- (6b)

   In this case, no "special" inter-router link is required for mobility
   -- the normal Internet connectivity between SBR1 and SBR2 would
   suffice.  However, it is quite likely that some sort of tunnelled
   link would still be desirable to offer protection of the VM image as
   it migrates.

6.4.  ILNP Capability in the Remote Host for VM Image Mobility



   For the remote host -- the CN -- the availability of ILNP would be
   beneficial.  However, for the first two scenarios listed above, as
   the packet signature of the transport flows remains fixed from the



Atkinson & Bhatti             Experimental                     [Page 29]

RFC 6748                        ILNP ADV                   November 2012


   viewpoint of the CN, it seems possible that the benefits of ILNP VM
   mobility could be used for datacentres even while CNs remain as
   normal IP hosts.  Of course, a major caveat here is that the
   application level protocols should be "well behaved": that is, the
   application protocol or configuration should not rely on the use of
   IP Addresses.

7.  Location Privacy



   Extending the Locator rewriting paradigm, it is possible to also
   enable Location privacy for ILNP by a modified version of the "onion
   routing" paradigm that is used for Tor [DMS04] [RSG98].

7.1.  Locator Rewriting Relay (LRR)



   To enable this function, we use a middlebox that we call the Locator
   Rewriting Relay.  The function of this unit is described by the use
   of Figure 7.1.

      <UDP: I_H, I_CN, P_H, P_CN><ILNP: L_1, L_CN>         --- (7a)

              v
              |
           +--+--+
           |     |   src=[I_H, L_1], L_X                   --- (7b)
           | LRR |   dst=[I_H, L_X], L_1                   --- (7c)
           |     |
           +--+--+
              |
              v
      <UDP: I_H, I_CN, P_H, P_CN><ILNP: L_X, L_CN>         --- (7d)

        LRR = Locator Rewriting Relay

     Figure 7.1: Locator Rewriting Relay (LRR) Example

   The operation of the LRR is conceptually very simple.  We assume that
   the LRR first has mappings as given in expressions (7b) and (7c) (see
   next subsection).  Expression (7b) says that for packets with src
   IL-V [I_H, L_1], the packet's source Locator value should be
   rewritten to value L_X and then forwarded.  Expression (7c) has the
   complimentary mapping for packets with destination IL-V [I_H, L_1]
   (for the reverse direction).

   Expression (6a) is a UDP/ILNP packet as might be sent in Figure 2.1
   from H to CN.  However, instead of going directly to L_CN, the packet
   with destination Locator L_1 goes to a LRR.  Expression (7d) is the
   result of the mapping of packet (7a) using expression (7b).



Atkinson & Bhatti             Experimental                     [Page 30]

RFC 6748                        ILNP ADV                   November 2012


   Note that it is entirely possible that the packet of expression (7d)
   then is processed by another LRR for source Locator value L_X.
   Effectively, this creates and LRR path for the packet, as an overlay
   path on top of the normal IP routing.

   In this way, there is a level of protection, without the need for
   cryptographic techniques, for the (topological) Location of the
   packet.  Of course, an extremely well-resourced adversary could,
   potentially, backtrack the LRR path, but, depending on the LRR
   overlay path that is created, could be very difficult to trace in
   reality.  For example, the mechanism will protect against off-path
   attacks, but where the threat regime includes the potential for on-
   path attacks, cryptographically protected tunnels between H and LRR
   might be required.

   Again, as the Locator value is not part of the end-to-end state, this
   mechanism is very general and has a low overhead.

7.2.  Options for Installing LRR Packet Forwarding State



   There are many options for managing the "network" of LRRs that could
   be in place if such a system was used on a large scale, including the
   setting up and removal of LRR state for packet relaying, as for
   expressions (7b) and (7c).  We consider this function to be outside
   the scope of these ILNP specifications, but note that there are many
   existing mechanisms that could modified for use, and also many
   possibilities for new mechanisms that would be specific to the use of
   ILNP LRRs.

   (Note also that the control/management communication with the LRR
   does not need to use ILNP: IPv4 or IPv6 could be used.)

   The host, H, by itself could install the required state, assuming it
   was aware of suitable information to contact the LRR.  The first
   packet in an ILNP session might contain a header option called a
   Locator Redirection Option (LRO).  The LRO would contain the Locator
   value that should be rewritten into the source Locator of the packet.
   When a LRR receives such a packet, it would install the required
   state.  Such a mechanism could be soft-state, requiring periodic use
   of the LRO in order to maintain the state in the LRR.  The LRO could
   also be delivered using an ICMP ECHO packet sent from H to the LRR,
   periodically, again to maintain a soft-state update.

   It would, of course, be prudent to protect the LRR state control
   packets with some sort of authentication token, to prevent an
   adversary from easily installing false LRR state and causing packets





Atkinson & Bhatti             Experimental                     [Page 31]

RFC 6748                        ILNP ADV                   November 2012


   from H or its correspondent to be subject to man-in-the-middle
   attacks, or black-holing.  Again, such attacks are not specific to
   ILNP or new to ILNP.

   It would also be possible to use proprietary application level
   protocols, with strong authentication for the control of the LRR
   state.  For example, an application level protocol based on XMPP
   (http://xmpp.org/) operating over SSL.

   Above, we have offered very brief and incomplete descriptions of some
   possibilities, and we do not necessarily mandate any one of them:
   they serve only as examples.

8.  Identity Privacy



   For the sake of completeness, and in complement to Section 6, it
   should be noted that ILNP can use either cryptographically verifiable
   Identifier values, or use Identifier values that provide a level of
   anonymity to protect a user's privacy.  More details are given in
   Sections 2 and 11 of [RFC6741].

9.  Security Considerations



   The relevant security considerations to this document are the same as
   for the main ILNP Architectural Description [RFC6740].  The one
   additional point to note is that this document describes ILNP
   capability in the SBR and so those adversaries wishing to subvert the
   operation of ILNP specifically, have a target that would,
   potentially, disable an entire site.  However, this is not an attack
   vector that is specific to ILNP: today, disruption of an IPv4 or IPv6
   SBR would have the same impact.

   The security considerations for Section 7 (Location Privacy) are
   already documented in [DMS04] and [RSG98].  One possibility is that
   the LRR mechanism itself could be used by an adversary to launch an
   attack and hide his own (topological) Location, for example.  This is
   already possible for IPv4 and IPv4 with a Tor-like system today, so
   is not new to ILNP.













Atkinson & Bhatti             Experimental                     [Page 32]

RFC 6748                        ILNP ADV                   November 2012


10.  References



10.1.  Normative 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.

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

   [RFC3022]     Srisuresh, P. and K. Egevang, "Traditional IP Network
                 Address Translator (Traditional NAT)", RFC 3022,
                 January 2001.

   [RFC3484]     Draves, R., "Default Address Selection for Internet
                 Protocol version 6 (IPv6)", RFC 3484, February 2003.

   [RFC4193]     Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast
                 Addresses", RFC 4193, October 2005.

   [RFC4632]     Fuller, V. and T. Li, "Classless Inter-domain Routing
                 (CIDR): The Internet Address Assignment and Aggregation
                 Plan", BCP 122, RFC 4632, August 2006.

   [RFC4787]     Audet, F., Ed., and C. Jennings, "Network Address
                 Translation (NAT) Behavioral Requirements for Unicast
                 UDP", BCP 127, RFC 4787, January 2007.

   [RFC4864]     Van de Velde, G., Hain, T., Droms, R., Carpenter, B.,
                 and E. Klein, "Local Network Protection for IPv6", RFC
                 4864, May 2007.

   [RFC4924]     Aboba, B., Ed., and E. Davies, "Reflections on Internet
                 Transparency", RFC 4924, July 2007.

   [RFC4984]     Meyer, D., Ed., Zhang, L., Ed., and K. Fall, Ed.,
                 "Report from the IAB Workshop on Routing and
                 Addressing", RFC 4984, September 2007.

   [RFC5902]     Thaler, D., Zhang, L., and G. Lebovitz, "IAB Thoughts
                 on IPv6 Network Address Translation", RFC 5902, July
                 2010.

   [RFC6177]     Narten, T., Huston, G., and L. Roberts, "IPv6 Address
                 Assignment to End Sites", BCP 157, RFC 6177, March
                 2011.




Atkinson & Bhatti             Experimental                     [Page 33]

RFC 6748                        ILNP ADV                   November 2012


   [RFC6740]     Atkinson, R. and S. Bhatti, "Identifier-Locator Network
                 Protocol (ILNP) Architectural Description", RFC 6740,
                 November 2012.

   [RFC6741]     Atkinson, R. and S. Bhatti, "Identifier-Locator Network
                 Protocol (ILNP) Engineering and Implementation
                 Considerations", RFC 6741, November 2012.

   [RFC6742]     Atkinson, R., Bhatti, S. and S. Rose, "DNS Resource
                 Records for the Identifier-Locator Network Protocol
                 (ILNP)", RFC 6742, November 2012.

   [RFC6743]     Atkinson, R. and S. Bhatti, "ICMPv6 Locator Update
                 Message", RFC 6743, November 2012.

   [RFC6744]     Atkinson, R. and S. Bhatti, "IPv6 Nonce Destination
                 Option for the Identifier-Locator Network Protocol for
                 IPv6 (ILNPv6)", RFC 6744, November 2012.

   [RFC6745]     Atkinson, R. and S. Bhatti,  "ICMP Locator Update
                 Message for the Identifier-Locator Network Protocol for
                 IPv4 (ILNPv4)", RFC 6745, November 2012.

   [RFC6746]     Atkinson, R. and S.Bhatti, "IPv4 Options for the
                 Identifier-Locator Network Protocol (ILNP)", RFC 6746,
                 November 2012.

   [RFC6747]     Atkinson, R. and S. Bhatti, "Address Resolution
                 Protocol (ARP) Extension for the Identifier-Locator
                 Network Protocol for IPv4 (ILNPv4)", RFC 6747, November
                 2012.

10.2.  Informative References



   [ABH07a]      Atkinson, R., Bhatti, S., and S. Hailes, "Mobility as
                 an Integrated Service Through the Use of Naming",
                 Proceedings of ACM Workshop on Mobility in the Evolving
                 Internet Architecture (MobiArch), ACM SIGCOMM, Kyoto,
                 Japan. 27 Aug 2007.

   [ABH07b]      Atkinson, R., Bhatti, S., and S. Hailes, "A Proposal
                 for Unifying Mobility with Multi-Homing, NAT, &
                 Security", Proceedings of 2nd ACM Workshop on Mobility
                 Management and Wireless Access (MobiWAC), ACM, Chania,
                 Crete, Oct 2007.  ISBN: 978-1-59593-809-1






Atkinson & Bhatti             Experimental                     [Page 34]

RFC 6748                        ILNP ADV                   November 2012


   [ABH08a]      Atkinson, R., Bhatti, S., and S. Hailes, "Mobility
                 Through Naming: Impact on DNS", Proceedings of 3rd ACM
                 Workshop on Mobility in the Evolving Internet
                 Architecture (MobiArch), ACM SIGCOMM, Seattle, WA, USA.
                 Aug 2008.

   [ABH08b]      Atkinson, R., Bhatti, S., and S. Hailes, "Harmonised
                 Resilience, Security, and Mobility Capability for IP",
                 Proceedings of the IEEE Military Communications
                 Conference (MILCOM), IEEE, San Diego, CA, USA, Nov
                 2008.

   [ABH09a]      Atkinson, R, Bhatti, S., and S. Hailes, "Site-
                 Controlled Secure Multi-Homing and Traffic Engineering
                 For IP", Proceedings of IEEE Military Communications
                 Conference (MILCOM), IEEE, Boston, MA, USA, Oct 2009.

   [ABH09b]      Atkinson, R., Bhatti, S., and S. Hailes, "ILNP:
                 Mobility, Multi-Homing, Localised Addressing and
                 Security Through Naming"", Telecommunication Systems",
                 vol. 42, no. 3-4, pp 273-291, Springer-Verlag, Dec
                 2009.

   [ABH10]       Atkinson, R., Bhatti, S., and S. Hailes, "Evolving the
                 Internet Architecture Through Naming", IEEE Journal on
                 Selected Areas in Communication (JSAC), vol. 28, no. 8,
                 pp 1319-1325, IEEE, Oct 2010.

   [appDNS]      Peterson, J., Kolkman, O., Tschofenig, H., and  B.
                 Aboba, "Architectural Considerations on Application
                 Features in the DNS", Work in Progress, July 2012.

   [BA11]        Bhatti, S. and R. Atkinson, "Reducing DNS Caching",
                 Proceedings of IEEE Global Internet Symposium (GI2011),
                 Shanghai, P.R. China, 15 Apr 2011.

   [BA12]        Bhatti, S. and R. Atkinson, "Secure & Agile Wide-area
                 Virtual Machine Mobility", Proceedings of IEEE Military
                 Communications Conference (MILCOM), Orlando, FL, USA,
                 Oct 2012.

   [BAK11]       Bhatti, S., Atkinson, R., and J. Klemets, "Integrating
                 Challenged Networks", Proceedings of IEEE Military
                 Communications Conference (MILCOM), IEEE, Baltimore,
                 MD, USA, Nov 2011.

   [BRDP11]      Boot, T. and A. Holtzer, "BRDP Framework", Work in
                 Progress, January 2011.



Atkinson & Bhatti             Experimental                     [Page 35]

RFC 6748                        ILNP ADV                   November 2012


   [DMS04]       Dingledine, R., Mathewson, N., and P. Syverson, "Tor:
                 the second-generation onion router", Proceedings of
                 13th USENIX Security Symposium, USENIX Association, San
                 Diego, CA, USA, 2004.

   [IEEE04]      "IEEE 802.1D - IEEE Standard for Local and Metropolitan
                 Area Networks, Media Access Control (MAC) Bridges",
                 IEEE Standards Association, New York, NY, USA, 9 June
                 2004.  Print: ISBN 0-7381-3881-5 SH95213.  PDF: ISBN
                 0-7381-3982-3 SS95213.

   [LABH06]      Atkinson, R., Lad, M., Bhatti, S., and S. Hailes, "A
                 Proposal for Coalition Networking in Dynamic
                 Operational Environments", Proceedings of IEEE Military
                 Communications Conference (MILCOM), IEEE, Washington,
                 DC, USA, Nov 2006.

   [mDNS11]      Cheshire, S. and M. Krochmal, "Multicast DNS", Work in
                 Progress, December 2011.

   [RAB09]       Rehunathan, D., Atkinson, R., and S. Bhatti, "Enabling
                 Mobile Networks Through Secure Naming", Proceedings of
                 IEEE Military Communications Conference (MILCOM), IEEE,
                 Boston, MA, USA, Oct 2009.

   [RB10]        Rehunathan, D. and S. Bhatti, "A Comparative Assessment
                 of Routing for Mobile Networks", Proceedings of 6th
                 IEEE International Conference on Wireless and Mobile
                 Computing Networking and Communications (WiMob), IEEE,
                 Niagara Falls, ON, Canada, Oct 2010.

   [RFC4193]     Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast
                 Addresses", RFC 4193, October 2005.

   [RFC6296]     Wasserman, M. and F. Baker, "IPv6-to-IPv6 Network
                 Prefix Translation", RFC 6296, June 2011.

   [RSG98]       Reed, M., Syverson, P., and D. Goldschlag, "Anonymous
                 Connections and Onion Routing", IEEE Journal on
                 Selected Areas in Communications, Vol. 16, No. 4, IEEE,
                 Piscataway, NJ, USA, May 1998.










Atkinson & Bhatti             Experimental                     [Page 36]

RFC 6748                        ILNP ADV                   November 2012


11.  Acknowledgements



   Steve Blake, Stephane Bortzmeyer, Mohamed Boucadair, Noel Chiappa,
   Wes George, Steve Hailes, Joel Halpern, Mark Handley, Volker Hilt,
   Paul Jakma, Dae-Young Kim, Tony Li, Yakov Rehkter, Bruce Simpson,
   Robin Whittle, and John Wroclawski (in alphabetical order) provided
   review and feedback on earlier versions of this document.  Steve
   Blake provided an especially thorough review of an early version of
   the entire ILNP document set, which was extremely helpful.  We also
   wish to thank the anonymous reviewers of the various ILNP papers for
   their feedback.

   Roy Arends provided expert guidance on technical and procedural
   aspects of DNS issues.

Authors' Addresses



   RJ Atkinson
   Consultant
   San Jose, CA 95125
   USA

   EMail: rja.lists@gmail.com


   SN Bhatti
   School of Computer Science
   University of St Andrews
   North Haugh, St Andrews
   Fife  KY16 9SX
   Scotland, UK

   EMail: saleem@cs.st-andrews.ac.uk


















Atkinson & Bhatti             Experimental                     [Page 37]