[ppml] Version think... was: alternative to 2005-1

[Kevin Loch]

Houle, Joseph D (Joe), CMO wrote:
> Problem: Multi-homing
> In IPv4, multi-advertising (i.e., advertising the same address across
> multiple providers) was the answer to the problem.
> In IPv6, shouldn't we be thinking of multi-addressing (i.e. multiple
> addresses for the same sub-net) as the answer?  ... Granted Shim6 is not
> fully baked yet, but that (or something which accomplishes the same
> basic goal) is where our creative energies should be focused."

A new new routing architectureis needed if we ever hope to fully utilize
the IPv6 address space.  500 Million /32's for LIR's is impossible to
imagine with todays architecture.

Fortunately there are only about ~21K AS's in use publicly.  If we
can optimize policy towards one prefix per AS it will help buy time
to develop a new architecture.

> Problem: ISP lock-in
> In IPv4, PI was the insurance against "my ISP got run over by a truck".
> In IPv6, shouldn't the prudent user (enterprise customer, campus, etc.)
> get addresses from all their service providers?

Each additional provider increases the chances of a renumbering event.
Despite the original promises, IPv6 still requires just as much manual
configuration as IPv4.

> We need to focus some attention on the effect of fragmentation (whether
> caused by PI or multi-advertising in general) on the Internet routing
> table size.   

Fortunately the large address size of IPv6 has enabled generous minimum
assignment sizes.  This will hopefully drive the prefix to AS ratio
close to 1.  In fact that is currently the case (IPv4 = ~8:1, IPv6 =
~1.25:1).

The proposed PI policies follow this generous assignment principle so
PI should not affect the prefix to AS ratio.  A significant benefit of
getting PI space in a separate address block is that filtering of
deaggregates becomes less painful.

We need a compatible migration path for most IPv4 address space holders,
and 2005-1 offers that.

- Kevin