Transport level multi-homing a structural view.

[Greg Maxwell <gmaxwell@martin.fl.us>]

We must first establish whether or not multihoming belongs above the
network layer.

I argue that:

IP is about transporting packets. The address does not just define a
host, it also defines the path to reach the host from a common root.

The network can not scale if we pollute the network with
unaggregateable routes. IPv6 has even more potential for pollution then
IPv4 if handled the same way. Only a relatively small percentage of edge
networks need to be multihomed to seriously pollute the route tables.

There is no reasonable way to globally multihome at the network layer
without polluting the global routing table. A simple order of magnitude
reduction (if even possible) is not even sufficent, future growth and
other capabilities (such as anycast, multicast, etc) may need the router
capacity. 

Multi-homing at a higher layer (such as in the transport) increases
flexibility, reduces network state, and improves end-to-endness.

*Changing to end-to-end multihoming after IP multi-homing is established
will be virtually impossible because of the massive structural
differences, to an even worse extent then unaggregated IPv4 routes.*

Because of these reasons, all long-term survivable global multi-homing
solutions must be implimented above the network layer and pollution must
be prevented from the start.

How might it look, network wise:

    [Tier 1 - provider & TLA][Tier 1 provider & TLA]
     |                      *
    [Tier 1 - provider & TLA][Tier 1 provider & TLA] - [Tier 1 ...]

(an arbitrary mesh of tier 1 providers,  
 route table size is their coustomers +  
 number of links to other tier 1s        
 routing capacity scales with links)     

    [Tier 1 - providerB & TLA][Tier 1 providerC & TLA]------------------\
     |                      *                                           |
    [Tier 1 - providerA & TLA][Tier 1 providerD & TLA] - [Tier 1 E ...] |
    |      |                    |   |                        |  |   |   |
    \     [tier 2 A prefix ]\   |   | /----------------------|  |   |   |
    [cust sngl homed A]	|  ||	|   [Tier 2 - D&E prefixes]     |   |   |    
    [multihomed cust A&D]--uu---/    | \[Cust D&E]   [Cust E]---/   |   |
    [cust sngl A]----------/|        |    /[Link Sharing cust Pre E&C]  |
                [Tier 3 provider A&D&E]   \[Link Sharing cust Pre C&E]--/
     [Cust A&D&E]/ \[Cust A&D&E]-----private IP level peering--/

Notice how this is so structurally differnt from today's network layer
multihomed network. Also notice the ability of customers to link share
without any provider asstiance, they just need to give each other some of
their address space. Tier >1 providers don't multihome.... They pass their
multihoming on to their customers.


--
The comments and opinions expressed herein are those of the author of this 
message and may not reflect the policies of the Martin County Board of
County Commissioners.