[83266] in North American Network Operators' Group
RE: IPv6 Address Planning
daemon@ATHENA.MIT.EDU (Cody Lerum)
Tue Aug 9 15:05:50 2005
Date: Tue, 9 Aug 2005 13:05:22 -0600
From: "Cody Lerum" <clerum@transaria.com>
To: "James" <james@towardex.com>
Cc: <nanog@merit.edu>
Errors-To: owner-nanog@merit.edu
Makes sense. However the PTP addresses need to be internally visible
from an NMS perspective in our network.
-C
-----Original Message-----
From: James [mailto:james@towardex.com]=20
Sent: Tuesday, August 09, 2005 12:13 PM
To: Cody Lerum
Cc: nanog@merit.edu
Subject: Re: IPv6 Address Planning
On Tue, Aug 09, 2005 at 11:24:22AM -0600, Cody Lerum wrote:
>=20
> Currently we are in the process of planning our IPv6 addressing schema
> for our network. We are a service provider with around 20 core=20
> routers, and several hundred enterprise customers. These customers=20
> currently connect back to our core via a separate VLANs or channelized
> DS1/DS3/OC-X type interfaces. Thus currently lots of /30 IPv4 blocks.
>=20
> Our address allocation is 2001:1940::/32
>=20
> Here is our current plan, but we are looking for suggestions from=20
> people who have been down this road before. The plan is to break out a
> /48 for our organization. Then break out the first /64 for loopbacks,=20
> and the next /64 for point-to-point connections. The PTP /64 then=20
> breaks out further into 1 /80 for core links, and 1 /80 for each of=20
> our distribution sites. Within these /80's are individual /112's for=20
> PTP links. What this will allow us to do is aggregate each sites PTP=20
> connections into /80's within our IGP.
The way we do it currently are as follows:
Reserve a /48 for backbone pointopoints (eg. 2001:4830:ff::/48) in US,
fe::/48 in EU. Reserve a /48 for loopbacks, and use /128s for each
loopback out of that. As for point to point links, we currently use
simple /64 subnets for each point to point (i.e. 2001:4830:ff:1500::/64,
etc where ::1 and ::2 are routers on either side of the circuit).
From there, we also have a /48 allocated per each POP for transfer
networks at that location for peering via pni and customer hand-offs.
Each xfer net is broken off as /64 out of that /48. We currently do not
perform any PTP link aggregation in our IGP, we simply ensure only
passive-interfaces are announced to IGP, thus PTP links are not even
present in the IGP table (only loopbacks and xfer nets/bgp next-hops
are).
It is not perfect but works well currently and scales just fine for us.
<shameless plug>
You may also find the ipv6-ops list helpful for v6 rollout discussions:
http://lists.cluenet.de/mailman/listinfo/ipv6-ops
</shameless plug>
James
--
James Jun
Infrastructure and Technology Services
TowardEX Technologies
Office +1-617-459-4051 x179 | Mobile +1-978-394-2867 james@towardex.com
| www.towardex.com
