Network Working Group                                           J. Arkko
Internet-Draft                                                  Ericsson
Expires: July 5, 2005                                       January 2005


     Failure Detection and Locator Selection Design Considerations
                 draft-ietf-shim6-failure-detection-00

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

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

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

   This Internet-Draft will expire on July 5, 2005.

Copyright Notice

   Copyright (C) The Internet Society (2005).

Abstract

   This draft discusses locator pair selection and failure detection
   mechanisms for the IPv6 multihoming feature being developed in the
   SHIM6 working group.  Elements of this document may also be useful
   for developing the details of the MOBIKE or HIP multihoming
   mechanisms.  The draft also discusses the roles of a multihoming
   protocol versus network attachment functions at IP and link layers.






Arkko                     Expires July 5, 2005                  [Page 1]


Internet-Draft           SHIM6 Failure Detection            January 2005


Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Related Work . . . . . . . . . . . . . . . . . . . . . . . . .  4
   3.  Definitions  . . . . . . . . . . . . . . . . . . . . . . . . .  7
         3.1   Available Addresses  . . . . . . . . . . . . . . . . .  7
         3.2   Locally Operational Addresses  . . . . . . . . . . . .  8
         3.3   Operational Address Pairs  . . . . . . . . . . . . . .  8
         3.4   Primary Address Pair . . . . . . . . . . . . . . . . . 10
         3.5   Miscellaneous  . . . . . . . . . . . . . . . . . . . . 10
   4.  Architectural Considerations . . . . . . . . . . . . . . . . . 11
   5.  An Approach  . . . . . . . . . . . . . . . . . . . . . . . . . 13
         5.1   State Machine for Addresses  . . . . . . . . . . . . . 13
         5.2   State Machine for Address Pair Selection . . . . . . . 14
         5.3   Pair Selection Algorithm . . . . . . . . . . . . . . . 18
         5.4   Protocol for Testing Unidirectional Reachability . . . 19
   6.  Security Considerations  . . . . . . . . . . . . . . . . . . . 22
   7.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 23
         7.1   Normative References . . . . . . . . . . . . . . . . . 23
         7.2   Informative References . . . . . . . . . . . . . . . . 23
       Author's Address . . . . . . . . . . . . . . . . . . . . . . . 25
   A.  Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 26
   B.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 27
       Intellectual Property and Copyright Statements . . . . . . . . 28



























Arkko                     Expires July 5, 2005                  [Page 2]


Internet-Draft           SHIM6 Failure Detection            January 2005


1.  Introduction

   The SHIM6 working group is extending IPv6 to support multihoming.  A
   number of possible approaches exist in this space, but the current
   focus of the group is to look at an IP layer (or layer 3.5) mechanism
   that hides multihoming from applications.  This mechanism needs to
   detect when a switch to another address or addresses becomes
   necessary.  We call this failure detection, because the SHIM6
   protocol works primarily as a failover rather than a load balancing
   scheme.

   This draft discusses what requirements such a component of the SHIM6
   protocol has, and how these requirements can be achieved.  The draft
   is structured as follows: Section 2 discusses what kind of solutions
   have been used in other similar protocols.  Section 3 defines a set
   of useful terms and discusses them, and Section 4 discusses the
   architectural implications of multihoming at IP layer.  Finally,
   Section 5 describes one possible solution involving two state
   machines, a failure testing protocol, and an address pair selection
   algorithm.

   For the purposes of this draft, we consider an address to be
   synonymous with a locator.  There may be other, higher level
   identifiers such as security associations, FQDNs, CGA public keys, or
   HITs that tie the different locators used by a node together.


























Arkko                     Expires July 5, 2005                  [Page 3]


Internet-Draft           SHIM6 Failure Detection            January 2005


2.  Related Work

   In SCTP [10], the addresses of the endpoints are learned in the
   connection setup phase either through listing them explictly or via
   giving a DNS name that points to them.  In order to provide a
   failover mechanism between multihomed hosts, SCTP has the following
   functions:


   o  One of the peer's addresses is selected as the primary address by
      the application running on top of SCTP.  All data packets are sent
      to this address until there is a reason to choose another address,
      such as the failure of the primary address.


   o  Testing the reachability of the peer endpoint's addresses.  This
      is done both via observing the data packets sent to the peer or
      via a periodic heartbeat when there is no data packets to send.

      Each time data packet retransmission is initiated (or when a
      heartbeat is not answered within the estimated round-trip time) an
      error counter is incremented.  When a configured error limit is
      reached, the particular destination address is marked as inactive.
      The reception of an acknowledgement or heartbeat response clears
      the counter.


   o  Retransmission: When retransmitting the endpoint attempts pick the
      most "divergent" source-destination pair from the original source-
      destination pair to which the packet was transmitted.  Rules for
      such selection are, however, left as implementation decisions in
      SCTP.

   SCTP does not define how local knowledge (such as information learned
   from the link layer) should be used.  SCTP also has no mechanism to
   deal with dynamic changes to the set of available addresses, although
   mechanisms for that are being developed [17].

   The MOBIKE protocol is currently being designed [15] [14].  This
   protocol operates in a mixed IPv4/IPv6 enviroment, and typically has
   to work through NATs.  The current design is assumed to need to work
   only in symmetric connectivity scenarios.

   Some of the issues that have been discussed in the MOBIKE design
   phase include the following:






Arkko                     Expires July 5, 2005                  [Page 4]


Internet-Draft           SHIM6 Failure Detection            January 2005


   o  Single address vs. multiple peer addresses.  A simple approach is
      to have the peers be aware of just the current address of the
      other side instead of all possible ones.  Assuming that one of the
      peers will request the other to start sending to a new address
      this works well.  However, this approach is unable to deal with
      problems that affect both nodes.  For instance, two nodes
      connected by two separate point-to-point links will be unable to
      switch to the other link if a failure occurs on the first one.


   o  Addresses vs. address pairs.  Are tests and current paths
      individual peer addresses, or pairs of peer and own addresses
      (paths)?  It seems that some failure scenarios require the use of
      a path rather than a single address.  A network failure may make
      it impossible to communicate between a particular pair of
      addresses, even if those addresses have some other connectivity.


   o  Where the connectivity information comes from.  Does it come from
      local stack (such as interface up/down, router advertisement),
      from reception of ESP packets, from IKEv2 keepalives, or through
      some MOBIKE-defined mechanism?

   The mobility and multihoming specification for the HIP protocol [13]
   leaves the determination of when address updates are sent to a local
   policy, but suggests the use of local information and ICMP error
   messages.

   Network attachment procedures are also relevant for multihoming.  The
   IPv6 and MIP6 working groups have standardized mechanisms to learn
   about networks that a node has attached to.  Basic IPv6 Neighbor
   Discovery was, however, designed primarily for static situations.
   The fully dynamic detection procedure has turned out to be a
   relatively complex procedure for mobile hosts, and it was not fully
   anticipated at the time IPv6 Neighbor Discovery or DHCP were being
   designed.  As a result, enhanced or optimized mechanisms are being
   designed in the DHC and DNA working groups [6] [7].

   ICE [16], STUN [11], and TURN [28] are also related mechanisms.  They
   are primarily used for NAT detection and communication through NATs
   in IPv4 environment, for application such as as voice over IP.  STUN
   uses a server in the Internet to discover the presence and type of
   NATs and the client's public IP addresses and ports.  TURN makes it
   possible to receive incoming connections in hosts behind NATs.  ICE
   makes use of these protocols in peer-to-peer cooperative fashion,
   allowing participants to discover, create and verify mutual
   connectivity, and then use this connectivity for multimedia streams.
   While these mechanisms are not designed for dynamic and failure



Arkko                     Expires July 5, 2005                  [Page 5]


Internet-Draft           SHIM6 Failure Detection            January 2005


   situations, they have many of the same requirements for the
   exploration of connectivity, as well as the requirement to deal with
   middleboxes.

   Related work in the IPv6 area includes RFC 3484 [5] which defines
   source and destination address selection rules for IPv6 in situations
   where multiple candidate address pairs exist.  RFC 3484 considers
   only a static situation, however, and does not take into account the
   effect of failures.  In the MULTI6 working group [23] considers how
   applications can re-initiate connections after failures in the best
   way.  This work differs from the shim-layer approach selected for
   further development in the working group with respect to the timing
   of the address selection.  In the shim-layer approach failure
   detection and the selection of new addresses happens at any time,
   while [23] considers only the case when an application re-establishes
   connections.



































Arkko                     Expires July 5, 2005                  [Page 6]


Internet-Draft           SHIM6 Failure Detection            January 2005


3.  Definitions

   This section defines terms useful in discussing the failure detection
   problem space.

3.1  Available Addresses

   SHIM6 nodes need to be aware of what addresses they themselves have.
   If a node loses the address it is currently using for communications,
   another address must replace this address.  And if a node loses an
   address that the node's peer knows about, the peer must be informed.
   Similarly, when a node acquires a new address it may generally wish
   the peer to know about it.

   Definition.  Available address.  An address is said to be available
   if the following conditions are fulfilled:


   o  The address has been assigned to an interface of the node.


   o  If the address is an IPv6 address, we additionally require that
      (a) the address is valid in the sense of RFC 2461 [2], and that
      (b) the address is not tentative in the sense of RFC 2462 [3].  In
      other words, the address assignment is complete so that
      communications can be started.

      Note this explicitly allows an address to be optimistic in the
      sense of [8] even though implementations are probably better off
      using other addresses as long as there is an alternative.


   o  The address is a global unicast, unique local address [9], or an
      unambiquous IPv6 link-local or IPv4 RFC 1918 address.  That is, it
      is not an IPv6 site-local address.  Where IPv6 link-local or RFC
      1918 addresses are used, their use needs to be unambiquous.  The
      precise meaning of ambiquous has not been defined yet, but one
      approach is requiring that at most one link-local address be used
      per node within the same connection between two peers.

         Note: Given RFC 3484 [5] rules for preferring smallest scope,
         it is likely that many IPv6 flows at least start with even
         link-local addresses.


   o  The address and interface is acceptable for use according to a
      local policy.




Arkko                     Expires July 5, 2005                  [Page 7]


Internet-Draft           SHIM6 Failure Detection            January 2005


   Available addresses are discovered and monitored through mechanisms
   outside the scope of SHIM6 (and HIP or MOBIKE).  These mechanisms
   include IPv6 Neighbor Discovery and Address Autoconfiguration [2]
   [3], DHCP [4], enhanced network detection mechanisms detected by the
   DNA working group, and corresponding IPv4 mechanisms, such as [6].

3.2  Locally Operational Addresses

   Two different granularity levels are needed for failure detection.
   The coarser granularity is for individual addresses:

   Definition.  Locally Operational Address.  An available address is
   said to be locally operational when its use is known to be possible
   locally: the interface is up and a relevant default router (if
   applicable) is known to be reachable.

   Locally operational addresses are discovered and monitored through
   mechanisms outside SHIM6 (and HIP or MOBIKE).  These mechanisms
   include IPv6 Neighbor Discovery [2], corresponding IPv4 mechanisms,
   and link layer specific mechanisms.

   Theoretically, it is also possible for hosts to learn about routing
   failures for a particular selected source prefix, even if no protocol
   exists today to distribute this information in a convenient manner.
   The development of such protocols would be possible, however.  One
   approach is overloading information in current IPv6 Router
   Advertisements (see [23]) or adding some new information in them.
   Similarly, hosts could learn information from servers that query the
   BGP routing tables [23].

3.3  Operational Address Pairs

   The existence of locally operational addresses are not, however, a
   guarantee that communications can be established with the peer.  A
   failure in the routing infrastructure can prevent the sent packets
   from reaching their destination.  For this reason we need the
   definition of a second level of granularity, for pairs of addresses:

   Definition.  Bidirectionally operational address pair.  A pair of
   locally operational addresses are said to be an operational address
   pair, iff bidirectional connectivity can be shown between the
   addresses.  That is, a packet sent with one of the addresses in the
   source field and the other in the destination field reaches the
   destination, and vice versa.

   Unfortunately, there are scenarios where bidirectionally operational
   address pairs do not exist.  For instance, ingress filtering or
   network failures may result in one address pair being operational in



Arkko                     Expires July 5, 2005                  [Page 8]


Internet-Draft           SHIM6 Failure Detection            January 2005


   one direction while another one is operational from the other
   direction.  The following definition captures this general situation:

   Definition.  Undirectionally operational address pair.  A pair of
   locally operational addresses are said to be an unidirectionally
   operational address pair, iff packets sent with the first address as
   the source and the second address as the destination can be shown to
   reach the destination.

   Both types of operational pairs are discovered and monitored through
   the following mechanisms:


   o  Positive feedback from upper layer protocols.  For instance, TCP
      can indicate to the IP layer that it is making progress.  This is
      similar to how IPv6 Neighbor Unreachability Detection can in some
      cases be avoided when upper layers provide information about
      bidirectional connectivity [2].  In the case of unidirectional
      connectivity, the upper layer protocol responses come back using
      another address pair, but show that the messages sent using the
      first address pair have been received.


   o  Negative feedback from upper layer protocols.  It is conceivable
      that upper layer protocols give an indication of a problem to the
      SHIM6 layer.  For instance, TCP could indicate that there's either
      congestion or lack of connectivity in the path because it is not
      getting ACKs.


   o  Explicit reachability tests.  For instance, the IKEv2 keepalive
      mechanism can be used to test that the current pair of addresses
      is operational.


   o  ICMP error messages.  Given the ease of spoofing ICMP messages,
      one should be careful to not trust these blindly, however.  Our
      suggestion is to use ICMP error messages only as a hint to perform
      an explicit reachability test, but not as a reason to disrupt
      ongoing communications without other indications of problems.  The
      situation may be different when certain verifications of the ICMP
      messages are being performed [22].  These verifications can ensure
      that (pratically) only on-path attackers can spoof the messages.
      Such verifications are not possible for all transport protocols,
      however.

   Note that some protocols, such as HIP [13], perform a return
   routability test of an address before it is taken into use.  The



Arkko                     Expires July 5, 2005                  [Page 9]


Internet-Draft           SHIM6 Failure Detection            January 2005


   purpose of this test is to ensure that fraudulent peers do not trick
   others into redirecting traffic streams onto innocent victims [31].
   Such tests can at the same time work as a means to ensure that an
   address pair is operational.  Note, however, that some advanced
   optimizations attempt to postpone the reachability tests so that they
   do not increase movement-related latency [29].

3.4  Primary Address Pair

   Contrary to SCTP which has a specific congestion avoidance design
   suitable for multi-homing, IP-layer solutions need to avoid sending
   packets concurrently over multiple paths; TCP behaves rather poorly
   in such circumstances.  For this reason it is necessary to choose a
   particular pair of addresses as the primary address pair which is
   used until problems occur, at least for the same session.

   A primary address pair need not be operational at all times.  If
   there is no traffic to send, we may not know if the primary address
   pair is operational.  Neverthless, it makes sense to assume that the
   address pair that worked in some time ago continues to work for new
   communications as well.

3.5  Miscellaneous

   Addresses can become deprecated [2].  When other operational
   addresses exist, nodes generally wish to move their communications
   away from the deprecated addresses.

   Similarly, IPv6 source address selection [5] may guide the selection
   of a particular source address - destination address pair.





















Arkko                     Expires July 5, 2005                 [Page 10]


Internet-Draft           SHIM6 Failure Detection            January 2005


4.  Architectural Considerations

   Architecturally, a number of questions arises.  One simple question
   is whether there needs to be communications between a multihoming
   solution residing at the IP layer and upper layer protocols?  Upon
   changing to a new address pair, transport layer protocol SHOULD be
   notified so that it can perform a slow start, or some other form of
   adaptation to the possibly changed conditions.  This is necessary,
   for instance, when switching from a high-bandwidth LAN interface to a
   low bandwidth cellular interface.  (Note that this notification can
   not be done in protocol designs where the end points are not the
   final hosts, such as where a gateway is used.)

   A more fundamental question is which protocols should be responsible
   for which parts of the problem.  It seems clear that no multihoming
   solution should take on the task of lower layers and other IP
   functions for discovering its own addresses or testing local
   connectivity.  Protocols such as DHCP or Neighbor and Router
   Discovery do this already.

   But it is less clear which protocol(s) should discover end-to-end
   connectivity problems or recover from them.  One answer is that this
   is clearly within the domain of multihoming protocol.  By performing
   testing and failure detection of the used path and switching to a new
   path if necessary, the transport and application protocols can work
   unchanged.

   On the other hand, one could argue that transport and application
   protocols would have more knowledge about the situation, and have a
   better ability to decide when a move is required.  For instance, they
   know what the required throughput and congestion status is.  Also, it
   would be unfortunate if both the IP layer and transport/application
   layer took action for the same problem, for instance by switching to
   a new address at the IP layer and throttling back due to "congestion"
   at the transport layer.

   One can also envision that applications would be able to tell the IP
   or transport layer that the current connection in unsatisfactory and
   an exploration for a better one would be desirable.  This would
   require an API to be developed, however.

   Generally speaking, we can divide information that a host has into
   three categories: local information from "lower layers" such as IPv6
   Neighbor Discovery, transit and congestion condition information from
   either from the multihoming protocol itself or from transport layer
   protocols and (where available) ECN, and application layer policies
   that dictate what the requirements are for acceptable connections.




Arkko                     Expires July 5, 2005                 [Page 11]


Internet-Draft           SHIM6 Failure Detection            January 2005


   The division of work is largely left as an open issue as far as this
   document is concerned, but our description works from a point of view
   of a multihoming protocol at the IP layer.  We also note that in the
   CELP proposal [18], both IP, transport, and application layer
   entities could share their connectivity status in a common
   information pool.  This may also be a useful approach.

   Finally, the last architectural question is about the difference
   between mobility and multihoming.  Given our definitions above,
   there's no fundamental difference with respect to how the
   multihoming/mobility protocol learns the addresses it has available.
   However, a practical difference is that in a multihoming scenario
   there are alternative addresses, whereas in mobility changes to a new
   address are forced due to the old address no longer being available.





































Arkko                     Expires July 5, 2005                 [Page 12]


Internet-Draft           SHIM6 Failure Detection            January 2005


5.  An Approach

   One suggested approach consists of a mechanism for keeping track of
   the host's own available addresses, operational addresses, and
   operational address pairs.

5.1  State Machine for Addresses

   Addresses can be in the AVAILABLE and OPERATIONAL states.  The state
   transitions relating to this are shown in Figure 1.

                     +--------------+
     Address becomes |              |
     available       |              |
   ----------------->|              |
                     |  AVAILABLE   |
   <-----------------|              |
     Address is no   |              |
    longer available |              |
                     +--------------+
                        |       / \
                Address |        | Address
                becomes |        | is no longer
            operational |        | operational
                        |        |
                       \ /       |
                     +--------------+
                     |              |
     Address is no   |              |
    longer available |              |
   <-----------------| OPERATIONAL  |
                     |              |
                     |              |
                     |              |
                     +--------------+

          Figure 1. Address state machine.


   When an address becomes operational, it SHOULD be reported as a new
   address to the peer.  Similarly, when an address is no longer
   operational or available, the peer SHOULD be informed.

   In addition, a particular address can be either preferred or
   deprecated.  This is not shown in the state machine.






Arkko                     Expires July 5, 2005                 [Page 13]


Internet-Draft           SHIM6 Failure Detection            January 2005


5.2  State Machine for Address Pair Selection

   A node runs the address pair selection state machine to choose the
   currently used primary address pair, the one which is used for
   sending outgoing packets.  A node runs one of these state machines
   towards each different peer, tracking the known address pairs and
   their status.  Each peer also has its own state machine for talking
   back to the node; there is no guarantee that the same address pairs
   (in reverse order) have the same state; lack of bidirectionally
   operational pair would result in a different state on both sides, for
   instance.

   The state machine can be in the NO PRIMARY, TESTING PRIMARY, and
   PRIMARY OPERATIONAL states.  The chosen address pair is known to be
   operational in the PRIMARY OPERATIONAL state, and is either
   unverified or non-operational in the other states.

   Figure 2 shows the state machine:

































Arkko                     Expires July 5, 2005                 [Page 14]


Internet-Draft           SHIM6 Failure Detection            January 2005


                         +----------------+
                         |                |
                         |                |
                         |                |
                         |                |
                         |       NO       |
                         |     PRIMARY    |
                         |                |
                   +-----|                |<---------------+
                   |     |                |                |
                   |     +----------------+                |
                   |         / \    / \                    |
               Add |          |      |                     |
             pair: |   Delete |      | Test         Delete |
              Send |   pair & |      | fail &       pair & |
              test |     Last |      | Last           Last |
                   |          |      |                     |
                   |     +----------------+                |
                   |     |                |                |
                   +---->|                |<----+          |
                         |                |     | Test     |
    Connect: Send test   |                |     | fail &   |
   --------------------->|     TESTING    |     | !Last    |
                         |     PRIMARY    |+----+          |
          +------------->|                |                |
          |              |                |<----+          |
          |        +---->|                |     |          |
          |        |     +----------------+     |          |
   Policy | ICMP | |          |      |          |          |
   change | Timer: |      ULP |      | Test     | Delete   |
          |   Send | feedback:|      | OK:      | pair &   |
          |   test |    Reset |      | Reset    | !Last    |
          |        |    timer |      | timer    |          |
          |        |         \ /    \ /         |          |
          |        |     +----------------+     |          |
          |        +-----|                |     |          |
          |              |                |-----+          |
          +--------------|                |                |
                         |                |                |
                   +-----|   OPERATIONAL  |                |
     ULP feedback: |     |     PRIMARY    |                |
       Reset timer |     |                |----------------+
                   +---->|                |
                         |                |
                         +----------------+

          Figure 2. Pair selection state machine.




Arkko                     Expires July 5, 2005                 [Page 15]


Internet-Draft           SHIM6 Failure Detection            January 2005


   The notation used in Figure 2 is explained below:


   Connect

      An event representing the desire of the application to send a
      packet to a new peer, or an indication from a peer wishing to
      connect to us.


   Test OK

      An event representing a successful completion of the reachability
      test.


   Test fail

      An event representing failure to complete the reachability test.


   ULP feedback

      An event representing positive indication from an upper layer
      protocol that the packets we have sent to the peer are getting
      through.


   ICMP

      An event representing the reception of an ICMP error message.


   Timer

      An event representing timer elapsing.


   Add pair

      An event representing the addition of a new possible address pair,
      either through learning a new local address or being told of a new
      remote address.  Note that this does not usually result in any
      immediate action, unless we are currently lacking an operational
      primary pair.






Arkko                     Expires July 5, 2005                 [Page 16]


Internet-Draft           SHIM6 Failure Detection            January 2005


   Delete pair

      An event representing the deletion of the currently chosen primary
      address pair.


   Policy change

      An event representing the desire of the local or remote end to
      change to a different address pair, despite the current one being
      operational.  This can be due to the availability of the higher-
      bandwidth connection, cost, or other issues.


   Last

      A condition that tells whether or not the currently chosen primary
      pair is the only known address pair.


   Send test

      An action to initiate the reachability test for a particular pair.
      This test is typically embedded in the SHIM6 connection setup
      exchange when run initially, and a separate exchange later.

      Note that due to potentially asymmetric connectivity, both sides
      have to perform their own tests, and make their own primary pair
      selections.


   Reset timer

      An action to reset a timer so that it will send an event after a
      specified time.

   The state machines also assumes an underlying multihoming signaling
   capabability, consisting of the following abstract message exchanges:


   Open

      Establishes a connection between the peers.  May also exchange
      locator sets and test reachability at the same time.







Arkko                     Expires July 5, 2005                 [Page 17]


Internet-Draft           SHIM6 Failure Detection            January 2005


   Test

      Verifies reachability using a specific address pair.


   Add

      Informs the peer about new locators.


   Delete

      Informs the peer about losing some locators.

   Note that the above state machine leaves open how specific address
   pairs are chosen, as this will be discussed in the next section.  We
   have also, on purpose, decided to avoid attaching functional labels
   such as "backup" to other address pairs beyond the primary pair.  It
   is our belief that a general design does not need these labels.

5.3  Pair Selection Algorithm

   The pair selection state machine assumes an ability to pick primary
   and alternative address pairs.

   This process results in a combinatorial explosion when there are many
   addresses on both sides.  Do both sides track all possible
   combinations of addresses?  If a failure occurs, shall all
   combinations be tested before giving up?  Are such tests performed in
   parallel or in sequence, and what kind of backoff procedures should
   be applied?

   Our suggestion is that nodes MUST first consult RFC 3484 [5] Section
   4 rules to determine what combinations of addresses are legal from a
   local point of view, as this reduces the search space.  RFC 3484 also
   provides a priority ordering among different address pairs, making
   the search possibly faster.  Nodes SHOULD also use local information,
   such as known quality of service parameters or interface types to
   determine what addresses are preferred over others, and try pairs
   containing such addresses first.  In some cases we can also learn the
   peer's preferences through the multihoming protocol [13].


      Discussion note 1: It may also be possible to simulate preferences
      by choosing to not tell the peer about some (non-preferred)
      addresses.





Arkko                     Expires July 5, 2005                 [Page 18]


Internet-Draft           SHIM6 Failure Detection            January 2005


      Discussion note 2: The preferences may either be learned
      dynamically or be configured.  It is believed, however, that
      dynamic learning based purely on the SHIM6 protocol is too hard
      and not the task this layer should do.  Solutions where multiple
      protocols share their information in a common pool of locators
      could provide this information from transport protocols, however
      [18].

   The reception of packets from the peer with a given address pair is a
   good hint that the address pair works, particularly when these
   packets are authenticated multihoming protocol packets.  However, the
   reception of these packets alone is an insufficient reason to switch
   to a new address, as in an unidirectional connectivity case the
   return path may not work.

   One suggested good implementation strategy is to record the
   reachability test result (an on/off value) and multiply this by the
   age of the information.  This allows recently tested address pairs to
   be chosen before old ones.

   Out of the set of possible candidate address pairs, nodes SHOULD
   attempt a test through all of them, but MUST do this sequentially
   (based on an implementation-dependent priority order) and using an
   exponential back-off procedure.

   This sequantial process is necessary in order to avoid a "signaling
   storm" when an outage occurs (particularly for a complete site).
   However, it also limits the number of addresses that can in practice
   be used for multihoming, considering that transport and application
   layer protocols will fail if the switch to a new address pair takes
   too long.  For instance, we can assume that an initial timeout value
   is 0.1 seconds and there are four addresses on both sides.  Going
   through all sixteen address pairs and doubling the timeout value at
   every trial would take 3200 seconds!

   Finally, as has been noted in the context of MOBIKE, the existence of
   NATs can require that peers continuously monitor the operational
   status of address pairs, as otherwise NAT state related to a
   particular communication is lost, and the peer on the outer side of
   the NAT can no longer reach the peer inside the NAT.

5.4  Protocol for Testing Unidirectional Reachability

   Testing for reachability is not easy in an environment where
   unidirectional reachability is a possibility.  This is because the
   test of a single pair may not result in a working paths to send both
   the request and response packets.  The following protocol could be
   used to avoid this problem:



Arkko                     Expires July 5, 2005                 [Page 19]


Internet-Draft           SHIM6 Failure Detection            January 2005


    Peer A                                        Peer B
      |                                             |
      |  Poll 1 (src=A1, dst=B1)                    |
      |-------------------------------------------->|
      |                                             |
      |               Poll 2 (src=B1, dst=A1) OK: 1 |
      |        X------------------------------------|
      |                                             |
      |  Poll 3 (src=A2, dst=B1)                    |
      |------------------------------X              |
      |                                             |
      |          Poll 4 (src=B2, dst=A1) OK: 1      |
      |<--------------------------------------------|
      |                                             |
      |  Poll 5 (src=A1, dst=B1) OK: 4              |
      |-------------------------------------------->|
      |                                             |


   When B receives the first Poll message, it memorizes that it has
   gotten it.  The Poll message from B, however, is lost so A tries
   again with another pair.  This is lost too, but B continues its own
   testing process by sending its second Poll message, which is received
   by A. The messages carry identifiers, and a list of identifiers that
   were found messages the sender had itself successfully received
   earlier.

   In the end of the example case, A and B know that they have a working
   path from A to B using (A1, B1) and from B to A using (B2, A1).

   More generally, when A decides that it needs to test for
   connectivity, it will initiate a set of Poll messages, in sequence,
   until it gets a Poll message from B indicating that (a) B has
   received one of A's Poll messages and, obviously, (b) that B's Poll
   message is getting through.  B uses the same algorithm, but starts
   the process from the reception of the first Poll mesage from A.

   Note that this protocol can be implemented in different ways.  One
   approach is to rely on data packets, such as TCP payload packets and
   acknowledgements.  This method has the benefit that it likely passes
   easily through firewalls and other middleboxes.  One exception to
   this are stateful firewalls that wish to know what happened "earlier"
   in the connection, but it seems that such firewalls are fundamentally
   incompatible with multi-homing anyway.  One drawback of this method
   is, however, that the the number of available payload packets may not
   match the need in a situation where a lot of address pairs need to be
   explored.




Arkko                     Expires July 5, 2005                 [Page 20]


Internet-Draft           SHIM6 Failure Detection            January 2005


   Another approach is to have a completely separate protocol for the
   exploration.  This would need to be explicitly allowed in firewalls
   before it could be used.  On the other hand, then it would be very
   clear for the firewall administrators what they are letting through.















































Arkko                     Expires July 5, 2005                 [Page 21]


Internet-Draft           SHIM6 Failure Detection            January 2005


6.  Security Considerations

   Attackers may spoof various indications from lower layers and the
   network in an effort to confuse the peers about which addresses are
   or are not working.  For example, attackers may spoof ICMP error
   messages in an effort to cause the parties to move their traffic
   elsewhere or even to disconnect.  Attackers may also spoof
   information related to network attachments, router discovery, and
   address assignments in an effort to make the parties believe they
   have Internet connectivity when in reality they do not.

   This may cause use of non-preferred addresses or even denial-of-
   service.

   SHIM6 does not provide any protection of its own for indications from
   other parts of the protocol stack.  However, MOBIKE is resistant to
   incorrect information from these sources in the sense that it
   provides its own security for both the signaling of addressing
   information as well as actual payload data transmission.  Denial-of-
   service vulnerabilities remain, however.  Some aspects of these
   vulnerabilities can be mitigated through the use of techniques
   specific to the other parts of the stack, such as properly dealing
   with ICMP errors [22], link layer security, or the use of [12] to
   protect IPv6 Router and Neighbor Discovery.



























Arkko                     Expires July 5, 2005                 [Page 22]


Internet-Draft           SHIM6 Failure Detection            January 2005


7.  References

7.1  Normative References

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

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

   [3]  Thomson, S. and T. Narten, "IPv6 Stateless Address
        Autoconfiguration", RFC 2462, December 1998.

   [4]  Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., and M.
        Carney, "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)",
        RFC 3315, July 2003.

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

   [6]  Aboba, B., "Detection of Network Attachment (DNA) in IPv4",
        draft-ietf-dhc-dna-ipv4-08 (work in progress), July 2004.

   [7]  Choi, J., "Detecting Network Attachment in IPv6 Goals",
        draft-ietf-dna-goals-00 (work in progress), June 2004.

   [8]  Moore, N., "Optimistic Duplicate Address Detection for IPv6",
        draft-ietf-ipv6-optimistic-dad-01 (work in progress), June 2004.

   [9]  Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast
        Addresses", draft-ietf-ipv6-unique-local-addr-05 (work in
        progress), June 2004.

7.2  Informative References

   [10]  Stewart, R., Xie, Q., Morneault, K., Sharp, C., Schwarzbauer,
         H., Taylor, T., Rytina, I., Kalla, M., Zhang, L., and V.
         Paxson, "Stream Control Transmission Protocol", RFC 2960,
         October 2000.

   [11]  Rosenberg, J., Weinberger, J., Huitema, C., and R. Mahy, "STUN
         - Simple Traversal of User Datagram Protocol (UDP) Through
         Network Address Translators (NATs)", RFC 3489, March 2003.

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

   [13]  Nikander, P., "End-Host Mobility and Multi-Homing with Host



Arkko                     Expires July 5, 2005                 [Page 23]


Internet-Draft           SHIM6 Failure Detection            January 2005


         Identity Protocol", draft-ietf-hip-mm-00 (work in progress),
         October 2004.

   [14]  Kivinen, T., "Design of the MOBIKE protocol",
         draft-ietf-mobike-design-00 (work in progress), June 2004.

   [15]  Eronen, P., "IKEv2 Mobility and Multihoming Protocol (MOBIKE)",
         draft-ietf-mobike-protocol-00 (work in progress), June 2005.

   [16]  Rosenberg, J., "Interactive Connectivity Establishment (ICE): A
         Methodology for Network  Address Translator (NAT) Traversal for
         Multimedia Session Establishment Protocols",
         draft-ietf-mmusic-ice-02 (work in progress), July 2004.

   [17]  Stewart, R., "Stream Control Transmission Protocol (SCTP)
         Dynamic Address  Reconfiguration",
         draft-ietf-tsvwg-addip-sctp-10 (work in progress),
         January 2005.

   [18]  Crocker, D., "Framework for Common Endpoint Locator Pools",
         draft-crocker-celp-00 (work in progress), February 2004.

   [19]  Dupont, F., "Address Management for IKE version 2",
         draft-dupont-ikev2-addrmgmt-05 (work in progress), June 2004.

   [20]  Eronen, P., "Mobility Protocol Options for IKEv2 (MOPO-IKE)",
         draft-eronen-mobike-mopo-00 (work in progress), July 2004.

   [21]  Eronen, P. and H. Tschofenig, "Simple Mobility and Multihoming
         Extensions for IKEv2 (SMOBIKE)", draft-eronen-mobike-simple-00
         (work in progress), March 2004.

   [22]  Gont, F., "ICMP attacks against TCP",
         draft-gont-tcpm-icmp-attacks-00 (work in progress),
         August 2004.

   [23]  Huitema, C., "Address selection in multihomed environments",
         draft-huitema-multi6-addr-selection-00 (work in progress),
         October 2004.

   [24]  Kivinen, T., "MOBIKE protocol",
         draft-kivinen-mobike-protocol-00 (work in progress),
         March 2004.

   [25]  Nordmark, E., "Multihoming without IP Identifiers",
         draft-nordmark-multi6-noid-02 (work in progress), July 2004.

   [26]  Nordmark, E., "Multihoming using 64-bit Crypto-based IDs",



Arkko                     Expires July 5, 2005                 [Page 24]


Internet-Draft           SHIM6 Failure Detection            January 2005


         draft-nordmark-multi6-cb64-00 (work in progress),
         November 2003.

   [27]  Nordmark, E., "Strong Identity Multihoming using 128 bit
         Identifiers (SIM/CBID128)", draft-nordmark-multi6-sim-01 (work
         in progress), October 2003.

   [28]  Rosenberg, J., "Traversal Using Relay NAT (TURN)",
         draft-rosenberg-midcom-turn-05 (work in progress), July 2004.

   [29]  Vogt, C., Arkko, J., Bless, R., Doll, M., and T. Kuefner,
         "Credit-Based Authorization for Mobile IPv6 Early Binding
         Updates", draft-vogt-mipv6-credit-based-authorization-00 (work
         in progress), May 2004.

   [30]  Ylitalo, J., "Weak Identifier Multihoming Protocol (WIMP)",
         draft-ylitalo-multi6-wimp-01 (work in progress), July 2004.

   [31]  Aura, T., Roe, M., and J. Arkko, "Security of Internet Location
         Management", In Proceedings of the 18th Annual Computer
         Security Applications Conference, Las Vegas, Nevada, USA.,
         December 2002.


Author's Address

   Jari Arkko
   Ericsson
   Jorvas  02420
   Finland

   Email: jari.arkko@ericsson.com



















Arkko                     Expires July 5, 2005                 [Page 25]


Internet-Draft           SHIM6 Failure Detection            January 2005


Appendix A.  Contributors

   This draft attempts to summarize the thoughts and unpublished
   contributions of many people, including the MULTI6 WG design team
   members Marcelo Bagnulo Braun, Iljitsch van Beijnum, Erik Nordmark,
   Geoff Huston, Margaret Wasserman, and Jukka Ylitalo, the MOBIKE WG
   contributors Pasi Eronen, Tero Kivinen, Francis Dupont, Spencer
   Dawkins, and James Kempf, and my colleague Pekka Nikander at
   Ericsson.  This draft is also in debt to work done in the context of
   SCTP [10].

   The protocol design in Section 5.4 is due to Erik, Marcelo, and
   Iljitsch.






































Arkko                     Expires July 5, 2005                 [Page 26]


Internet-Draft           SHIM6 Failure Detection            January 2005


Appendix B.  Acknowledgements

   The author would also like to thank Christian Huitema, Pekka Savola,
   and Hannes Tschofenig for interesting discussions in this problem
   space, and for their comments on earlier versions of this draft.














































Arkko                     Expires July 5, 2005                 [Page 27]


Internet-Draft           SHIM6 Failure Detection            January 2005


Intellectual Property Statement

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.


Disclaimer of Validity

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
   ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED,
   INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
   INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.


Copyright Statement

   Copyright (C) The Internet Society (2005).  This document is subject
   to the rights, licenses and restrictions contained in BCP 78, and
   except as set forth therein, the authors retain all their rights.


Acknowledgment

   Funding for the RFC Editor function is currently provided by the
   Internet Society.




Arkko                     Expires July 5, 2005                 [Page 28]