Re: [v6ops] How to divide v6ops

[Tina TSOU <[e-mail]>]

Fred,
The questions we listed below were just temporary appear in some of the v4v6tran discussions.
I believe the v4v6tran motivation is pretty much in line with you, "operating the IPv4 side of a dual stack network", and transition to IPv6.

Here is the framework I-D for v4v6tran.
https://datatracker.ietf.org/doc/draft-ietf-v6ops-v4v6tran-framework/
Framework for IP Version Transition Scenarios

I don't think it is just *NOG issues.


Best Regards,
Tina TSOU
http://tinatsou.weebly.com/contact.html
http://www.huawei.com/en/solutions/expand-broadband/hw-092950-ipv6.htm



-----Original Message-----
From: [e-mail] [mailto:[e-mail]] On Behalf Of Fred Baker
Sent: Thursday, August 18, 2011 3:36 PM
To: Brian E Carpenter
Cc: IPv6 Operations; Ron Bonica
Subject: Re: [v6ops] How to divide v6ops


On Aug 18, 2011, at 3:06 PM, Brian E Carpenter wrote:

> Hi Fred,
> 
> Will this reduce the total workload and/or reduce cross-posting?
> I have some doubts, since it seems likely that there will be crossover 
> issues all the time. Nevertheless, dividing the workload does seem 
> like a reasonable principle. However, it's only when we see a full 
> list of the drafts to be assigned to the new WG that we can really see 
> if it will work out.
> 
> I don't like the name "v4ops". It seems somewhat contradictory to 
> draft-george-ipv6-support, although the proposed charter is completely 
> in line with that draft. Ideally, we'd call it "legacy-ip-ops" but 
> that isn't possible, so how about "v46ops"?

My thought, which I'm not stuck on but is what I'm thinking right now, is that this is actually not very much about anything with a "6" in it. I'm in favor of pushing IPv6 as a technology, but...

It's about the fact that networks that are deploying IPv6 still have to stay in business, and as a result have v4 networks that have to stay running until a certain percentage of their access customers have upgraded CPEs, edge networks, and hosts. Host upgrade is a limited timeframe problem, given that people buy new ones at some rate. CPEs etc can be another discussion; if I have a working CPE, why do I spend $$ to change it? It's in the ISP's interest for them to all go spend that money; I doubt that the average consumer will see it as in their interest also.

Since Tina brought it up, let me comment on how this differs from v4v6tran. It differs in that it is only partially about transition issues or explaining to operators that have completely ignored IPv6 for 15 years and gotten no education the issues brought up in v4v6tran:

   o  How to decide the IPv6 address architecture in the network?
   o  What is the recommended prefix length for a large operator?
   o  What is the recommended prefix length for a medium operator?
   o  What is the recommended prefix length to hand out to customers?
   o  What is the recommended longest prefix length an operator should
      accept from customers?
   o  If privacy is a concern and an operator wants to use ULA in the
      network, what are the guidelines?
   o  How to deploy an IPv4 access network over an IPv6 core network?
   o  How to deploy an IPv6 access network over an IPv4 core network?
   o  Under what considerations, IS-IS should be used?
   o  Under what considerations, OSPFv3 should be used?
   o  What is the longest prefix to be allowed for peering?

and so on. Those are, as the working group pointed out at IETF-79 with very few punches pulled, *NOG questions or "go take one of the many fine courses on the topic".

And it is not at all about introducing someone's latest and greatest tweak on how to put an IPv6 packet into a tunnel, put an IPv4 packet into a tunnel, or slice and dice a port number, unless we're talking about operational experience with deployed technologies.

It is about operating the IPv4 side of a dual stack network. CGN, a+p, and all that.

Attached I have a spreadsheet (csv format) in which I tried to classify where I thought existing documents might go. No comment on whether the working group should actually be working on the draft; my question was "if someone sent in this draft as a -00, which working group might they choose and why". I might obviously be wrong. But it's on the basis of this that I said that the breakout was in the neighborhood of even.