Internet Engineering Task Force                 Jieyun (Jessica) Yu
INTERNET DRAFT                                                UUNET
Expires February, 2000                                 August, 1999

draft-yu-simple-ipv6multihoming-00.txt


                    A Simple Solution for IPv6 Multihoming

Status of this Memo

   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026.

   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.

   Distribution of this memo is unlimited.

   Copyright Notice

   Copyright (C) The Internet Society (1999).  All Rights Reserved.

Abstract

   There is a challenge of providing IPv6 multihoming sites with the
   basic benefits such as redundancy and load sharing, and at the same
   time, scaling the global IPv6 routing table. It is equally
   challenging to provide a solution which is simple enough to be
   operational manageable. This document proposes a simple solution
   which meets the requirements and is also operationally manageable.

1. Purpose

   There is a challenge of providing IPv6 multihoming sites with the
   basic benefits such as redundancy and load sharing, and at the same
   time, scaling the global IPv6 routing table. It is equally

Yu              A Simple Solution for IPv6 Multihoming          [Page 1]


Internet Draft   draft-yu-simple-ipv6multihoming-00.txt      August 1999

   challenging to provide a solution which is simple enough to be
   operational manageable. This document proposes a routing and
   addressing scheme that supports IPv6 multihoming which achieves the
   followings:

      a. Providing redundancy and load sharing for the multi-homed sites
      b. Scaling the global IPv6 Internet routing table
      c. Simple and operationally manageable

2. Proposal

                           ISP-3 ---- ISP-4
                             |      /  |
                             |    /    |
                             |  /      |
                           ISP-1 ---- ISP-2
                    link-1   \          /  link-2
                              Customer-A

   A multi-homed customer will designate a primary ISP and receive IP
   address from its primary ISP's IPv6 aggregation block. In this
   example, ISP-1 is the primary ISP of customer-A and assigns Addr-1-A
   to the customer. Customer-A will advertise addr-1-A to ISP-1 and
   ISP-2 respectively. ISP-2 will advertise Addr-1-A to ISP-1 and to
   ISP-1 only. ISP-1 will, of course, advertise its own aggregation
   Addr-1 to the entire Internet.

   For inbound traffic to Customer-A, traffic will first be forwarded to
   ISP-1 since it advertises Addr-1 which includes Addr-1-A. ISP-1 will
   then forward the traffic destined to Customer-A via its connection to
   ISP-2 and/or its direct link to customer-A, according to certain
   polices. Most common policy would be the use of the shortest exit. By
   using both connections to forward traffic to Customer-A, load sharing
   is achieved.

   For outbound traffic from Customer-A, ISP-1 and ISP-2 can announce
   default and/or specific prefixes based on customer-A's requirements
   and traffic originated from Customer-A will be forwarded accordingly.

   When the link between customer-A and ISP-2 (link-2 in the figure)
   goes down, all traffic will go in/out via the connection between
   Customer-A and ISP-1 (link-1 in the figure). Likewise, when link-1 is
   experiencing an outage, link-2 will be used for the traffic.  This is
   because ISP-1 will continue announce its aggregate block (i.e. Addr-
   1) to the entire Internet and ISP-2 will still advertise Addr-1-A to
   ISP-1, all the inbound traffic to the customer will be take the path:
   ISP-1 -> ISP-2 -> Customer-A. Outbound traffic will automatically
   fall to link-2. This way, redundancy is achieved.

Yu              A Simple Solution for IPv6 Multihoming          [Page 2]


Internet Draft   draft-yu-simple-ipv6multihoming-00.txt      August 1999

   Same mechanism can be extended to those sites multihoming to more
   than two ISPs. Only those ISPs that the customer directly connected
   to would carry the more specific prefix assigned to the multi-homed
   customer.

   The more specific prefix announcement could be covered under the
   peering agreement between ISPs as stated in [RFC2546].

3. Advantages of the proposal

      - Achieves the goal of scaling the global routing table.

      - It scales the global routing table. Only ISPs involved in the
      multihoming attachment need to know the more specific route of
      Addr-1. Other ISPs such as ISP-3 and ISP-4 in figure 1 do not need
      to know specifics.

      If the two involved ISPs has no direct connection, the more
      specific route would need to be carried by other ISPs in the path.
      This seems to be lesser a problem since ISP assigned with an
      Internet visible aggregate block usually has direct connection to
      each other.

      - It is really simple and thus more manageable and scalable
      operation wise. It does not require multiple addresses assigned
      all individual hosts of a multi-homed customer site, neither it
      requires O(N^2) tunnels among different ISP routers.

4. Concerns

      - The primary ISP of a multihoming customer (such as ISP-1 in the
      example) would need to do more work in terms of distributing the
      traffic among other connected ISPs and the customer link.  This
      can be done using BGP policy (BGP community) and using shortest
      exit. It will also carry more traffic for the customer. However,
      this is a value added service to the customer and the primary ISP
      can charge more fees for such services.

      - The primary ISP is the sole interface for the multihomed
      customer to the Internet other than the ISPs the customer has
      directly connection with. Outages such as one between ISP-1 and
      ISP-4 would impact the reachability from customers of ISP-4 to
      Customer-A even though ISP-2 still has good connection to ISP-4.
      However, if the primary ISP is a good quality ISP, this sort of
      situation should happen rarely. The reason is that it's common
      practice that an ISP, especially good quality ISP, has multiple
      connections to other big ISPs at different geographical locations.
      Good quality ISPs also have robust network design to prevent any

Yu              A Simple Solution for IPv6 Multihoming          [Page 3]


Internet Draft   draft-yu-simple-ipv6multihoming-00.txt      August 1999

      failure to impact the whole network. Choosing a good quality and
      robust ISP as primary ISP is a good practice.

   It is desirable to have solution which provides perfect redundancy
   and, at the same time, simplicity to scale management and operation.
   In the absence of a perfect solution, the trade-off needs to be made.
   The author believes that it is very important that the solution has
   to be simple and manageable. Manageability should be the top
   consideration.

5. Security Considerations

   Security considerations are out of scope of this document.

6. References

   [RFC2546] A. Durand and B. Buclin, "6Bone Routing Practice."
   RFC2546, March 1999.ftp://ftp.isi.edu/in-notes/rfc2546.txt.

Author's Address

   Jieyun (Jessica) Yu
   UUNET Technology
   880 Technology Dr.
   Ann Arbor, MI 48108

   Phone: (734) 214-7314

   EMail: jyy@uu.net

Yu              A Simple Solution for IPv6 Multihoming          [Page 4]

------- End of Forwarded Message