Dual stack statistics?

Mike Jones mike at mikejones.in
Thu Feb 28 17:00:25 CET 2013

On 28 February 2013 07:10, Tassos Chatzithomaoglou
<achatz at forthnetgroup.gr> wrote:
> Mike Jones wrote on 28/02/2013 02:17:
>> On 27 February 2013 20:56, Tassos Chatzithomaoglou
>> <achatz at forthnetgroup.gr> wrote:
>>> I had a quick look at this month's accounting data from a bras (around 4k dual-stack users) and here are some more numbers:
>>> IPv4/IPv6 upload: 98%/2%
>>> IPv4/IPv6 download: 93%/7%
>>> Supposing the acct records are right (we've hit quite a few bugs on radius ipv6 attrib records), i find this quite low...
>> I agree that does seem quite a bit lower than the numbers other
>> networks are giving. Is this for residential services?
> Yes.
>> My first thought was 'how dual stacked' are they? ie. do you have
>> control of most of the CPE, or are they likely to be using their own?
> Either their own (largest percentage), or ours (after they manually enable IPv6).
>> That prompts me to ask the list in general - does anyone have any
>> numbers on what %age of customers with their own CPE push some kind of
>> v6 traffic over a dual stacked connection? The main focus seems to be
>> overall number of users and traffic levels, but I suspect this is also
>> an interesting deployment status metric, which will become more
>> relevant as more ISPs dual stack.
>> - Mike
> Can you be more specific with the term %age?

I know customer numbers are normally carefully released, but I expect
many people wouldn't mind sharing that out of 50,000 dual stack lines
sampled 10,000* did some kind of v6. I don't believe I have seen v4/v6
usage figures presented like this before.

I assume ISPs who provide their own CPE would see some v6 traffic from
>95% of customers, but that might also be an interesting 'problem'
area if anyone has looked in to it.

- Mike

*wishful thinking?

More information about the ipv6-ops mailing list