Re: [v6ops] discussion of transition technologies

[Lee Howard <[e-mail]>]

On 1/22/18, 3:05 PM, "Fred Baker" <[e-mail]> wrote:

>
>
>On Jan 22, 2018, at 11:54 AM, Lee Howard <[e-mail]> wrote:
>> On 1/22/18, 2:20 PM, "Ole Troan" <[e-mail]> wrote:
>>>>>> I think I might also argue that the market has more or less followed
>>>>>> that advice. Your spreadsheet seems to suggest that.
>>>>> 
>>>>> The interesting thing is that 6rd, which is a way of appearing to
>>>>>have
>>>>> an
>>>>> IPv6 network without actually having one, is not what one might call
>>>>> "prevalent". It has in fact been used for *transition*, in places
>>>>>like
>>>>> Free - which used to connect IPv6 customers using 6rd and (I
>>>>> understand)
>>>>> has recently announced native IPv6 deployment. The places I know that
>>>>> have used it used it for a while and then have gone native.
>>>>> 
>>>>> Would you agree with that?
>>>> 
>>>> I would; that is my perception. MHO is that 6rd has had its day, and
>>>> while
>>>> I don’t think it needs to be deprecated, I haven’t heard any scenarios
>>>> in
>>>> the past several years where it solves an actual problem.
>>> 
>>> Apart from giving millions of users IPv6 access?
>> 
>> Why would anyone want to do that?
>> 
>> That’s a bit of a snarky question, but it’s a real one. Is there any
>> real-world problem for which 6rd is the best answer?
>
>I think the problem space I'm familiar with is stated in my previous note
>in this thread (scroll up). Imagine that you are a network operator (ISP
>or enterprise) and have some IPv4-dependent system (or set of systems)
>that you haven't replaced yet (or can't replace) that you need to jump
>over to create an IPv6 deployment. That might be as simple as - what I
>understand to have been Free's business case until recently - you have an
>IPv4 network and want to rapidly deploy IPv6. Or, the case I'm told
>another company had, that its DSL infrastructure was IPv4-only and 6rd
>allowed it to factor that out of the discussion until they could fix it
>(now, I'm told, fixed). "IPv6 islands connected by IPv4 infrastructure",
>which might be continents and peninsulas, who knows.

Yes, I’m familiar with those use cases. They had edge networks that
couldn’t do IPv6. I haven’t seen provider edge equipment shipped in the
last ten years that is IPv6-incapable. I don’t think whether they had
business reasons for wanting IPv6.

Sander’s example is similar, but was L2 switching gear that couldn’t do
security inspections at L3, and he also says, “part of that has solved
itself over the years."

>
>Now, to my mind and I think yours, the best answer to "I want to deploy
>IPv6" is "so do so". Where it becomes an issue is "but that costs money
>for some reason". 6rd might be a temporary measure to provide the service
>until the money was no longer an issue.

I’m thinking like a business person: why do you *want* to deploy IPv6?

Or as an engineer: I have twenty projects I could work on, what am I
trading off to find that 6rd is the optimal choice?


>
>> “I can’t update my network to support IPv6, but there are IPv6-only
>>hosts
>> that my users need to be able to reach” is the scenario 6rd addresses.
>>Is
>> that an actual case? The case “My ISP hasn’t updated to support IPv6,
>>but
>> there are IPv6-only hosts I need to reach” is solved with a tunnel
>>broker.
>> 
>> I don’t deny that it is deployed at scale. I’m asking whether there are
>> any new deployments, recent or contemplated, and what path on a decision
>> tree would lead one to decide “6rd.”


I’m not trying to deprecate 6rd; it is in active use, and it has been good
for the growth of the protocol, to prove that there was interest.

What I’m struggling to believe is that there is an actual case now, where
somebody was considering how to deploy IPv6, and I would advise them to
use 6rd. Has anyone here encountered that in the last couple of years? If
you had the conversation now, would you still advise 6rd?

Lee