Mailing List Archive

On Tue, 26 Oct 2021, David Conrad wrote:

> Ah. Cogent. I suspect IPv6 peering policies. Somebody should bake a
> cake.

According to https://twitter.com/Benjojo12/status/1452673637606166536
Cogent<->Google IPv6 now works. A cake is in order, but perhaps a
celebratory one!?

--
Mikael Abrahamsson email: swmike@swm.pp.se

On 22/10/2021 17:08, tim@pelican.org wrote:
> I don't think it'll ever make money, but I think it will reduce
> costs. CGNAT boxes cost money, operating them costs money, dealing
> with the support fallout from them costs money. Especially in the
> residential space, where essentially if the customer calls you, ever,
> you just blew years' worth of margin.

There aren't enough folk thinking along these lines, so thank you for
writing it.

Every flow you can route exclusively with 6, is one flow you aren't
having to pay extra for so it can sit in a CGNAT state table.

... And that's before they call you, as Tim also rightly points out.

--
Tom

Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
traffic flow through CGNAT, so cheaper hardware and less IPv4 address
space. Isn't it?

22.10.21 20:19, Mark Tinka ????:
>
>
> On 10/22/21 18:08, tim@pelican.org wrote:
>
>> I don't think it'll ever make money, but I think it will reduce
>> costs.  CGNAT boxes cost money, operating them costs money, dealing
>> with the support fallout from them costs money.  Especially in the
>> residential space, where essentially if the customer calls you, ever,
>> you just blew years' worth of margin.
>
> The problem is accurately modelling cost reduction using native IPv6 in
> lieu of CG-NAT is hard when the folk that need convincing are the CFO's.
>
> They are more used to "spend 1 to get 2". Convincing them to "save 2 by
> spending 1" - not as easy as one may think.
>
> Mark.
>

On 11/3/21 22:13, Max Tulyev wrote:

> Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
> traffic flow through CGNAT, so cheaper hardware and less IPv4 address
> space. Isn't it?

How to express that in numbers CFO can take to the bank?

Mark.

With a kicking ass pitch

-----Original Message-----
From: NANOG <nanog-bounces+jean=ddostest.me@nanog.org> On Behalf Of Mark Tinka
Sent: November 26, 2021 5:52 AM
To: nanog@nanog.org
Subject: Re: IPv6 and CDN's



On 11/3/21 22:13, Max Tulyev wrote:

> Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
> traffic flow through CGNAT, so cheaper hardware and less IPv4 address
> space. Isn't it?

How to express that in numbers CFO can take to the bank?

Mark.

On 26/11/2021 13:52, Mark Tinka wrote:
> On 11/3/21 22:13, Max Tulyev wrote:
>> Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
>> traffic flow through CGNAT, so cheaper hardware and less IPv4 address
>> space. Isn't it?
>
> How to express that in numbers CFO can take to the bank?

"want to buy 5 of those shiny new CGNAT boxes or only 2 ?"

Frank

Well … YMMV. We’ve been running v6 for years, and it didn’t really make a dent in spend or boxes or rate of v4 depletion. Big part of the problem in our neck of the woods is millions of v4-only terminals … as well as large customer/gov bids requiring tons of v4 address space.

> On Nov 26, 2021, at 07:04, Jean St-Laurent via NANOG <nanog@nanog.org> wrote:
>
> ?With a kicking ass pitch
>
> -----Original Message-----
> From: NANOG <nanog-bounces+jean=ddostest.me@nanog.org> On Behalf Of Mark Tinka
> Sent: November 26, 2021 5:52 AM
> To: nanog@nanog.org
> Subject: Re: IPv6 and CDN's
>
>
>
>> On 11/3/21 22:13, Max Tulyev wrote:
>>
>> Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
>> traffic flow through CGNAT, so cheaper hardware and less IPv4 address
>> space. Isn't it?
>
> How to express that in numbers CFO can take to the bank?
>
> Mark.
>

Care to explain because the alternative seems pretty self-evident.




-----
Mike Hammett
Intelligent Computing Solutions
http://www.ics-il.com

Midwest-IX
http://www.midwest-ix.com

----- Original Message -----

From: "Jose Luis Rodriguez" <jlrodriguez@gmail.com>
To: "Jean St-Laurent" <jean@ddostest.me>
Cc: nanog@nanog.org
Sent: Friday, November 26, 2021 8:16:53 AM
Subject: Re: IPv6 and CDN's

Well … YMMV. We’ve been running v6 for years, and it didn’t really make a dent in spend or boxes or rate of v4 depletion. Big part of the problem in our neck of the woods is millions of v4-only terminals … as well as large customer/gov bids requiring tons of v4 address space.

> On Nov 26, 2021, at 07:04, Jean St-Laurent via NANOG <nanog@nanog.org> wrote:
>
> With a kicking ass pitch
>
> -----Original Message-----
> From: NANOG <nanog-bounces+jean=ddostest.me@nanog.org> On Behalf Of Mark Tinka
> Sent: November 26, 2021 5:52 AM
> To: nanog@nanog.org
> Subject: Re: IPv6 and CDN's
>
>
>
>> On 11/3/21 22:13, Max Tulyev wrote:
>>
>> Implementing IPv6 reduces costs for CGNAT. You will have (twice?) less
>> traffic flow through CGNAT, so cheaper hardware and less IPv4 address
>> space. Isn't it?
>
> How to express that in numbers CFO can take to the bank?
>
> Mark.
>

Here are some maths and 1 argument kicking ass pitch for CFO’s that use iphones.

Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4

https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/



Build around that maybe?



Jean



From: Mike Hammett <nanog@ics-il.net>
Sent: November 26, 2021 11:56 AM
To: Jose Luis Rodriguez <jlrodriguez@gmail.com>
Cc: nanog@nanog.org; Jean St-Laurent <jean@ddostest.me>
Subject: Re: IPv6 and CDN's



Care to explain because the alternative seems pretty self-evident.



-----
Mike Hammett
Intelligent Computing Solutions
http://www.ics-il.com

Midwest-IX
http://www.midwest-ix.com



_____

With that specific line directly from Apple:



"And when IPv6 is in use, the median connection setup is 1.4 times faster than IPv4. This is primarily due to reduced NAT usage and improved routing."



There it is, Improved routing.



Jean



From: Jean St-Laurent <jean@ddostest.me>
Sent: November 26, 2021 4:44 PM
To: 'Mike Hammett' <nanog@ics-il.net>; 'Jose Luis Rodriguez' <jlrodriguez@gmail.com>
Cc: 'nanog@nanog.org' <nanog@nanog.org>
Subject: RE: IPv6 and CDN's



Here are some maths and 1 argument kicking ass pitch for CFO’s that use iphones.

Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4

https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/



Build around that maybe?



Jean



From: Mike Hammett <nanog@ics-il.net <mailto:nanog@ics-il.net> >
Sent: November 26, 2021 11:56 AM
To: Jose Luis Rodriguez <jlrodriguez@gmail.com <mailto:jlrodriguez@gmail.com> >
Cc: nanog@nanog.org <mailto:nanog@nanog.org> ; Jean St-Laurent <jean@ddostest.me <mailto:jean@ddostest.me> >
Subject: Re: IPv6 and CDN's



Care to explain because the alternative seems pretty self-evident.



-----
Mike Hammett
Intelligent Computing Solutions
http://www.ics-il.com

Midwest-IX
http://www.midwest-ix.com



_____

On 11/26/21 1:44 PM, Jean St-Laurent via NANOG wrote:
>
> Here are some maths and 1 argument kicking ass pitch for CFO’s that
> use iphones.
>
> *Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4*
>
> https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/
>
> Build around that maybe?
>
>
This really hits my bs meter big time. I can't see how nat'ing is going
to cause a 40% performance hit during connections. The article also
mentions http2 (and later v3) which definitely make big improvements so
I'm suspecting that the author is conflating them.

Mike

On Fri, Nov 26, 2021 at 6:07 PM Michael Thomas <mike@mtcc.com> wrote:

>
> On 11/26/21 1:44 PM, Jean St-Laurent via NANOG wrote:
>
> Here are some maths and 1 argument kicking ass pitch for CFO’s that use
> iphones.
>
> *Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4*
>
>
> https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/
>
>
>
> Build around that maybe?
>
> This really hits my bs meter big time. I can't see how nat'ing is going to
> cause a 40% performance hit during connections. The article also mentions
> http2 (and later v3) which definitely make big improvements so I'm
> suspecting that the author is conflating them.
>
> Mike
>

Ok, take the same ipv6 is faster claim from facebook

https://www.internetsociety.org/blog/2015/04/facebook-news-feeds-load-20-40-faster-over-ipv6/



>

On 11/26/21 3:11 PM, Ca By wrote:
>
>
> On Fri, Nov 26, 2021 at 6:07 PM Michael Thomas <mike@mtcc.com> wrote:
>
>
> On 11/26/21 1:44 PM, Jean St-Laurent via NANOG wrote:
>>
>> Here are some maths and 1 argument kicking ass pitch for CFO’s
>> that use iphones.
>>
>> *Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4*
>>
>> https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/
>>
>> Build around that maybe?
>>
>>
> This really hits my bs meter big time. I can't see how nat'ing is
> going to cause a 40% performance hit during connections. The
> article also mentions http2 (and later v3) which definitely make
> big improvements so I'm suspecting that the author is conflating them.
>
> Mike
>
>
> Ok, take the same ipv6 is faster claim from facebook
>
> https://www.internetsociety.org/blog/2015/04/facebook-news-feeds-load-20-40-faster-over-ipv6/
>
>
Still really thin with details of why. At least this says that they are
NAT'ing v4 at *their* edge. But 99% of the lag of filling your newsfeed
is their backend and transport, not connection times so who knows what
they are actually measuring. Most NAT'ing is done at the consumer end by
your home router in any case.

Mike

Might also be due to Happy Eyeballs 2 artificial IPv4 A resolution delay.

https://www.researchgate.net/publication/343467289_Reducing_User_Perceived_Latency_in_Smart_Phones_Exploiting_IP_Network_Diversity/fulltext/5f2b6bdaa6fdcccc43ac7b99/Reducing-User-Perceived-Latency-in-Smart-Phones-Exploiting-IP-Network-Diversity.pdf

I don’t know the current Apple IOS HE2 postpone delay.

JC Bisecco

De : NANOG <nanog-bounces+jc=jclb.net@nanog.org> De la part de Michael Thomas
Envoyé : samedi 27 novembre 2021 00:20
À : Ca By <cb.list6@gmail.com>
Cc : nanog@nanog.org
Objet : Re: IPv6 and CDN's



On 11/26/21 3:11 PM, Ca By wrote:


On Fri, Nov 26, 2021 at 6:07 PM Michael Thomas <mike@mtcc.com<mailto:mike@mtcc.com>> wrote:


On 11/26/21 1:44 PM, Jean St-Laurent via NANOG wrote:
Here are some maths and 1 argument kicking ass pitch for CFO’s that use iphones.
Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4
https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/

Build around that maybe?


This really hits my bs meter big time. I can't see how nat'ing is going to cause a 40% performance hit during connections. The article also mentions http2 (and later v3) which definitely make big improvements so I'm suspecting that the author is conflating them.

Mike

Ok, take the same ipv6 is faster claim from facebook

https://www.internetsociety.org/blog/2015/04/facebook-news-feeds-load-20-40-faster-over-ipv6/



Still really thin with details of why. At least this says that they are NAT'ing v4 at *their* edge. But 99% of the lag of filling your newsfeed is their backend and transport, not connection times so who knows what they are actually measuring. Most NAT'ing is done at the consumer end by your home router in any case.

Mike

We now have apple and fb saying ipv6 is faster than ipv4.



If we can onboard Amazon, Netflix, Google and some others, then it is a done deal that ipv6 is indeed faster than ipv4.



Hence, an easy argument to tell your CFO that you need IPv6 for your CDN.



Xmas is coming so the budget season. Who knows. You might get lucky this year.



From: NANOG <nanog-bounces+jean=ddostest.me@nanog.org> On Behalf Of Michael Thomas
Sent: November 26, 2021 6:20 PM
To: Ca By <cb.list6@gmail.com>
Cc: nanog@nanog.org
Subject: Re: IPv6 and CDN's





On 11/26/21 3:11 PM, Ca By wrote:





On Fri, Nov 26, 2021 at 6:07 PM Michael Thomas <mike@mtcc.com <mailto:mike@mtcc.com> > wrote:



On 11/26/21 1:44 PM, Jean St-Laurent via NANOG wrote:

Here are some maths and 1 argument kicking ass pitch for CFO’s that use iphones.

Apple tells app devs to use IPv6 as it's 1.4 times faster than IPv4

https://www.zdnet.com/article/apple-tells-app-devs-to-use-ipv6-as-its-1-4-times-faster-than-ipv4/



Build around that maybe?



This really hits my bs meter big time. I can't see how nat'ing is going to cause a 40% performance hit during connections. The article also mentions http2 (and later v3) which definitely make big improvements so I'm suspecting that the author is conflating them.

Mike



Ok, take the same ipv6 is faster claim from facebook



https://www.internetsociety.org/blog/2015/04/facebook-news-feeds-load-20-40-faster-over-ipv6/





Still really thin with details of why. At least this says that they are NAT'ing v4 at *their* edge. But 99% of the lag of filling your newsfeed is their backend and transport, not connection times so who knows what they are actually measuring. Most NAT'ing is done at the consumer end by your home router in any case.

Mike

On 11/26/21 4:15 PM, Jean St-Laurent wrote:
>
> We now have apple and fb saying ipv6 is faster than ipv4.
>
> If we can onboard Amazon, Netflix, Google and some others, then it is
> a done deal that ipv6 is indeed faster than ipv4.
>
> Hence, an easy argument to tell your CFO that you need IPv6 for your CDN.
>
Netflix is already v6 ready. The biggest obstacle is probably aws
because that's where a lot of the long tail of the internet resides.
Lobbying them would get the most bang for the buck.

Mike

AWS has been gradually improving support and adding features. They just
announced this service, which might help with adoption:

https://aws.amazon.com/about-aws/whats-new/2021/11/aws-nat64-dns64-communication-ipv6-ipv4-services/


On Fri., Nov. 26, 2021, 19:28 Michael Thomas, <mike@mtcc.com> wrote:

>
> On 11/26/21 4:15 PM, Jean St-Laurent wrote:
>
> We now have apple and fb saying ipv6 is faster than ipv4.
>
>
>
> If we can onboard Amazon, Netflix, Google and some others, then it is a
> done deal that ipv6 is indeed faster than ipv4.
>
>
>
> Hence, an easy argument to tell your CFO that you need IPv6 for your CDN.
>
> Netflix is already v6 ready. The biggest obstacle is probably aws because
> that's where a lot of the long tail of the internet resides. Lobbying them
> would get the most bang for the buck.
>
> Mike
>
>

On 11/26/21 4:30 PM, Oliver O'Boyle wrote:
> AWS has been gradually improving support and adding features. They
> just announced this service, which might help with adoption:
>
> https://aws.amazon.com/about-aws/whats-new/2021/11/aws-nat64-dns64-communication-ipv6-ipv4-services/

That's a start, I guess. Before all they had was some weird VPN
something or other. Let me guess though: they are monetizing their
market failure.

Mike


>
>
> On Fri., Nov. 26, 2021, 19:28 Michael Thomas, <mike@mtcc.com> wrote:
>
>
> On 11/26/21 4:15 PM, Jean St-Laurent wrote:
>>
>> We now have apple and fb saying ipv6 is faster than ipv4.
>>
>> If we can onboard Amazon, Netflix, Google and some others, then
>> it is a done deal that ipv6 is indeed faster than ipv4.
>>
>> Hence, an easy argument to tell your CFO that you need IPv6 for
>> your CDN.
>>
> Netflix is already v6 ready. The biggest obstacle is probably aws
> because that's where a lot of the long tail of the internet
> resides. Lobbying them would get the most bang for the buck.
>
> Mike
>
>

But CFOs like monetization. Was that thread about IPv6 or CFO?



From: Michael Thomas <mike@mtcc.com>
Sent: November 26, 2021 7:37 PM
To: Oliver O'Boyle <oliver.oboyle@gmail.com>
Cc: Jean St-Laurent <jean@ddostest.me>; Ca By <cb.list6@gmail.com>; North American Network Operators' Group <nanog@nanog.org>
Subject: Re: IPv6 and CDN's



That's a start, I guess. Before all they had was some weird VPN something or other. Let me guess though: they are monetizing their market failure.

On 11/26/21 4:39 PM, Jean St-Laurent wrote:
>
> But CFOs like monetization. Was that thread about IPv6 or CFO?
>

Amazon's in this case. They are monetizing their lack of v6 support
requiring you go through all kinds of expensive hoops instead of doing
the obvious and routing v6 packets.

Mike

> *From:*Michael Thomas <mike@mtcc.com>
> *Sent:* November 26, 2021 7:37 PM
> *To:* Oliver O'Boyle <oliver.oboyle@gmail.com>
> *Cc:* Jean St-Laurent <jean@ddostest.me>; Ca By <cb.list6@gmail.com>;
> North American Network Operators' Group <nanog@nanog.org>
> *Subject:* Re: IPv6 and CDN's
>
> That's a start, I guess. Before all they had was some weird VPN
> something or other. Let me guess though: they are monetizing their
> market failure.
>

On Fri., Nov. 26, 2021, 19:41 Michael Thomas, <mike@mtcc.com> wrote:

>
> On 11/26/21 4:39 PM, Jean St-Laurent wrote:
>
> But CFOs like monetization. Was that thread about IPv6 or CFO?
>
>
> Amazon's in this case. They are monetizing their lack of v6 support
> requiring you go through all kinds of expensive hoops instead of doing the
> obvious and routing v6 packets.
>

They're getting better at it, at least. They also recently added v6 support
in their NLBs and you can get a /56 for every VPC for direct access. I
don't think they offer BYO v6 yet, as they do for v4, but it will come.


Mike
>
>
>
> *From:* Michael Thomas <mike@mtcc.com> <mike@mtcc.com>
> *Sent:* November 26, 2021 7:37 PM
> *To:* Oliver O'Boyle <oliver.oboyle@gmail.com> <oliver.oboyle@gmail.com>
> *Cc:* Jean St-Laurent <jean@ddostest.me> <jean@ddostest.me>; Ca By
> <cb.list6@gmail.com> <cb.list6@gmail.com>; North American Network
> Operators' Group <nanog@nanog.org> <nanog@nanog.org>
> *Subject:* Re: IPv6 and CDN's
>
>
>
> That's a start, I guess. Before all they had was some weird VPN something
> or other. Let me guess though: they are monetizing their market failure.
>
>

> On Oct 26, 2021, at 9:11 AM, David Conrad <drc@virtualized.org> wrote:
>
> There has been some effort to create a governance model for the root server system (see https://www.icann.org/en/system/files/files/rssac-037-15jun18-en.pdf) but I believe it has gotten bogged down in the question of “what do you do when a root server operator isn’t doing the job ‘right’ (whatever that means and after figuring out who decides) but doesn’t want to give up being a root server operator?”.

Unless you actually read the document. The process is that the fact is recognized and documented, the Designation and Removal function advises the ICANN board, they adopt a resolution, and instruct the IANA to remove the addresses from the relevant files, and from that point on nobody NEW tries to usevtheRSO. If someone does ask the company a question, they might or might not respond, and it might even have correct data, but then again might not. We can’t control who sends us requests, and we don’t have a black list.

On 11/26/21 15:00, Jean St-Laurent wrote:

> With a kicking ass pitch

Can I take your CFO :-)...

Mark.

On 11/26/21 15:47, Frank Habicht wrote:

>
> "want to buy 5 of those shiny new CGNAT boxes or only 2 ?"

To which she will respond, "2 or 5, what do I make :-)?"

Mark.

On 11/26/21 16:16, Jose Luis Rodriguez wrote:

> Well … YMMV. We’ve been running v6 for years, and it didn’t really make a dent in spend or boxes or rate of v4 depletion. Big part of the problem in our neck of the woods is millions of v4-only terminals … as well as large customer/gov bids requiring tons of v4 address space.

I can very easily see why "IPv6 saves you on CG-NAT capex might not be
entirely true" in cases such as these.

On paper, it all adds up.

Mark.