Re: Heads up - Plea for allocating a /8 to ISPs

Le 11 nov. 2010 à 11:28, Joel Jaeggli a écrit :

> On 11/11/10 11:03 AM, Rémi Després wrote:
>> 
>> Le 11 nov. 2010 à 10:43, Brian E Carpenter a écrit :
>> 
>>>> I know that some believe that purposely breaking the IPv4 service could be a way to promote IPv6.
>>> 
>>> I don't believe that, and I don't doubt the good intentions of the
>>> people proposing this - I've seen too much of them in v6 working groups
>>> for that. The problem is elsewhere:
>> 
>>> this would be a very strong message
>>> to ISPs that are *not* actively planning v6 support that here is a way
>>> to evade the issue for many more years.
>> 
>> I don't see this at all.
>> No one can have doubt that I help ISPs to plan native IPv6 "service".
> 
> Ipv4 service is going to get worse. natting it at the carrier level, is
> inevitable as it is today in many of our enviroments. and as an operator
> I don't fundamentaly see the difference regarding what addresses are on
> the inside and outside of the nat, each one is an event horizon. some
> customers are going to experience breakage, in the form of their
> existing cpe failing at some point along the way.
> 
> As operators we have a fiduciary responsbility to our business, our
> customers and the oganizations of which we are member's from which we
> secure addresses, to:
> 
> A - use them judiciously
> B - secure the resource we need to support business growth
> 
> I happen to disagree that this prefix is necessary, not because I think
> that providers shouldn't use nat444 as one of the tools in their
> arsenal. I didn't support it in the rir's and I don't support it here.

Fair enough.
ISPs aren't all the same.

But you won't be hurt if some others use this new /8 prefix.

My point isn't that all ISPs  should use it.
It is only that those that ask for it have a valid reason to do so.
I se no good reason to refuse it. 
(The idea that this would demotivate them to offer also IPv6 seems to me insulting to their ability to understand the situation.)

> 
> It's possible that with sufficient thrust disagreement can be overcome,
> that's fine but the argumentation in favor of it either in the rir's or
> here has been strictly unconvincing as far as I'm concerned,  I don't
> belive that special use ipv4 blocks assigned at this late date are in
> the interests of existing or future ip address users.

My impression is that you negate that 10/8 used for both internal and external addresses creates problems with some NATs.
AFAIK, this is a sad fact, but a fact.

Regards,
RD




> We have enough
> cruft of that form in ipv4 already.
> 
>> IMHO, ISPs that ask for this prefix would rather interpret the message as:
>> "We IETF don't care about your view of your real ISP problems. We use to solve OUR problems, and wish to continue to do so."
>> I don't say, it is what you mean, I just say that this interpretation seems to me legitimate.
>> 
>> 
>>> So we end up with a bigger
>>> legacy of customers stuck behind NAT444 and ISPs that have no intention
>>> to move forward.
>>> 
>>> I would like to be proved wrong.
>> 
>> NAT444 customers are IMHO less a problem than those that prematurely use IPv6 with partially broken solutions.
>> 
>> ISP cooperation is needed to cleanly deploy IPv6.
>> Having respect for their legitimate requests seems to me the best path toward clean IPv6 everywhere.
>> (I don't think we will be stuck if we do what we have to.)
>> 
>> Regards,
>> RD
>> 
>>> 
>>> On 2010-11-11 15:30, Rémi Després wrote:
>>>> Hi, Ron,
>>>> 
>>>> 
>>>> Le 9 nov. 2010 à 13:26, Ronald Bonica a écrit :
>>>>> I am having a strange sense of déjà vu. About two years ago, we saw another proposal to allocate a /8 for private use to support another transition mechanism. We concluded:
>>>>> 
>>>>> - that we didn't have a /8 to spare
>>>> 
>>>> - Allocating a /8 to ISPs, while it is still feasible, would only slightly modify the date at which IPv4 prefixes er exhausted, not a big deal.
>>>> 
>>>>> - that a /8 wouldn't be enough, anyway
>>>> 
>>>> - ISPs already operate several parallel instances of 10/8 clouds, so that, although a /7 would be more generous than better a /8, a new /8 is enough to replace current 10/8's by a prefix private sites don't uses internally.
>>>> 
>>>> The known reason why ISPs have problems if they assign 10/8's is that some NAT44s:
>>>> . assign 10/8's for internal use
>>>> . cannot work with identical internal and external address spaces 
>>>> 
>>>> CONCLUSION:
>>>> Assigning a new /8 to ISPs so that they can replace 10/8:
>>>> - is easy to do
>>>> - has negligible effect on the transition schedule
>>>> - does solve a real problem
>>>> It should, in view of this analysis above, be highly recommended by IETF
>>>> 
>>>> I know that some believe that purposely breaking the IPv4 service could be a way to promote IPv6.
>>>> IMHO, their time would be better spent working to eliminate IPv6 limitations than militating against IPv4 deserved QoS.
>>>> 
>>>> Kind regards,
>>>> RD
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>>> So, the proposal was abandoned. I wonder if we are not in the exact same place today.
>>>>> 
>>>>> Maybe the equation would be changed if we could make do with a /16, but it appears that a /16 wouldn't help.
>>>>> 
>>>>>                                         Ron
>>>>>                                         <speaking as individual contributor>
>>>>> 
>>>>> BTW, the LISP WG is talking about asking for a prefix for a very similar reason.
>>>>> 
>>>>> 
>>>>>> -----Original Message-----
>>>>>> From: [email protected] [mailto:[email protected]] On Behalf
>>>>>> Of james woodyatt
>>>>>> Sent: Monday, November 08, 2010 11:50 PM
>>>>>> To: IPv6 Operations; [email protected]; [email protected]
>>>>>> Subject: Re: [v6ops] Heads up
>>>>>> 
>>>>>> On Nov 8, 2010, at 20:24 , Erik Kline wrote:
>>>>>>> Didn't Akira Nakagawa/KDDI present something like this a few meetings
>>>>>> ago?  I can't recall clearly...
>>>>>> 
>>>>>> It was in Chicago, if I recall correctly.  At which point Tony Hain
>>>>>> joked that we should direct IANA to just turn over the entire remaining
>>>>>> free pool for that use.
>>>>>> 
>>>>>> Personally, I'd prefer to take Tony's suggestion as a serious proposal.
>>>>>> A single /8 isn't enough, is it?  For NAT444 to be a really workable
>>>>>> architecture, we've got to give over enough addresses that service
>>>>>> providers can handle at least a hundred million subscriber addresses in
>>>>>> the same address realm, right?
>>>>>> 
>>>>>> 
>>>>>> --
>>>>>> james woodyatt <[email protected]>
>>>>>> member of technical staff, communications engineering
>>>>>> 
>>>>>> 
>>>>>> _______________________________________________
>>>>>> v6ops mailing list
>>>>>> [email protected]
>>>>>> https://www.ietf.org/mailman/listinfo/v6ops
>>>>> _______________________________________________
>>>>> v6ops mailing list
>>>>> [email protected]
>>>>> https://www.ietf.org/mailman/listinfo/v6ops
>>>> 
>>>> 
>>>> _______________________________________________
>>>> v6ops mailing list
>>>> [email protected]
>>>> https://www.ietf.org/mailman/listinfo/v6ops
>>>> 
>>> 
>> 
>> 
>> _______________________________________________
>> v6ops mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/v6ops
>> 
> 


_______________________________________________
v6ops mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/v6ops