Re: Comments on IPv6 Prefix Subdelegation
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Comments on IPv6 Prefix Subdelegation

Maybe you can help me reword it. What I am getting at is this:
a) within the home, in the example in figure 3, I have four routers and nine IP subnets. For each router to know where in the home to send data, the usual thing is for the routers in the home to do is communicate with the others using a routing protocol.
b) there are two CPE routers, each of which has an upstream router in its ISP. Since they are not exchanging a routing protocol with their upstream, they will need a static route in their own tables and to advertise a default upstream route within the home.
c) RFC 3704 observes on the ingress filtering performed in the upstream routers, and suggests that the two CPE routers should have some way to ensure that they only send traffic that will pass the filter to their upstream. Hence, each CPE Router might have a filter installed that looks at the source address of a datagram and when necessary forwards it to the other CPE. Or if we had source/ destination routing, could advertise the relevant prefix with its default route, so that the three routers (not CPE routers, just plain old routers, but probably with a firewall filter configured due to the observation about corporate information security policies applying to telecommuting home offices) would be able to send traffic to the right CPE.
How would you suggest I word this? In my mind, taking what is written there and confusing it with the relationship with the upstream ISP requires a strange reading of the text, which is all about routing within the home. 

On Jul 28, 2009, at 3:51 PM, Azinger, Marla wrote:
Fred: Here is the paragraph that is worded in a way that leads me to thinking you are saying to do OSPF to the upstream. I believe something needs to be taken out or added to clarify it: 

Routing in such an environment calls for a routing protocol such as
  RIPv6 [RFC2080], IS-IS [RFC5308], or OSPF [RFC5340].  In addition,
  each CPE router will need to install a static default route upstream
  and advertise a default route in the chosen routing protocol.  The
  issues raised in [RFC3704] also apply, meaning that the two CPE
  routers may each need to observe the source addresses in datagrams
  they handle to divert them to the other CPE to handle upstream

-----Original Message-----
From: Azinger, Marla
Sent: Tuesday, July 28, 2009 4:25 AM
To: 'Fred Baker'
Cc: IETF IPv6 Mailing List; draft-ietf-v6ops-ipv6-cpe-router at tools.ietf.org ; draft-donley-ipv6-cpe-rtr-use-cases-and-reqs at tools.ietf.org 
Subject: RE: Comments on IPv6 Prefix Subdelegation
Im thinking one step further than the double routers. For example if these routers are not serviced by something at least the service type of a dedicated T1 to each router then they would be doing VPN. So there are more requirements that need to be met here to make OSPF a realistic option. 

Thank you
Marla

-----Original Message-----
From: Fred Baker [mailto:fred at cisco.com]
Sent: Tuesday, July 28, 2009 2:59 AM
To: Azinger, Marla
Cc: IETF IPv6 Mailing List; draft-ietf-v6ops-ipv6-cpe-router at tools.ietf.org ; draft-donley-ipv6-cpe-rtr-use-cases-and-reqs at tools.ietf.org 
Subject: Re: Comments on IPv6 Prefix Subdelegation


On Jul 28, 2009, at 11:39 AM, Azinger, Marla wrote:


2.  I have concern regarding the suggestions in section 2.3   Am I
interpreting this correctly that you are suggesting upstreams do OSPF
over VPN with residential customers?


within their homes?
No, I am suggesting that in a home that has more than one router, one might want an IGP, just like one does in other places.

Note: Messages sent to this list are the opinions of the senders and do not imply endorsement by the IETF.