Addressing versus Routing (Was: Deploying IPv6 in a datacenter)

Michael.Dillon at btradianz.com Michael.Dillon at btradianz.com
Thu Dec 22 11:55:50 UTC 2005


> However I'm much more concerned that "big" providers (anyone who can
> qualify for a /32) need to make nearly zero changes to their way of 
> doing things, but Mom&Pop's regional ISP or Chuck's Web Hosting and 
> Bait Shop are going to be losing out big when it comes to IPv6. 
> Which is preferable, giving /32's to people who don't need anywhere 
> near that much space so that they can traffic engineer things they 
> way they need to, or being more flexible when it comes to 
> deaggregation when strictly necessary? 
> 
> Shim6 and others are interesting, and solve multihoming issues for 
> some, but they don't address traffic engineering or the need to do 
> more with smaller allocations.

This is why I have suggested that we need to open up 
additional IPv6 addresses for geo-topological addressing.
This means that instead of getting one big /32, you would
be able to apply for a seaparate allocation for each city
in whatever size is appropriate for each city. Then, you
can multihome inside that city and your announcements 
won't clutter up the global routing table because they
will be replaced by a single big city aggregate that covers
all the small and medium sized companies in the city.
The providers that offer such multihoming inside any
given city will need to interconnect inside that city
either at an IX or privately.

The RIRs have not made any decisions yet about offering
geotop addresses, but 7/8 of the IPv6 address space has
been reserved for different types of allocation schemes.
I believe that geotop addresses or something similar,
will eventually be offered by ARIN and the other RIRs.
In order to get to that point, small and mid-sized companies
need to make their voices heard.

--Michael Dillon




More information about the NANOG mailing list