measurement data of IPv6 flows and IPv6 download rate for Dual-Stack subscribers

[Tassos Chatzithomaoglou]

Thanks a lot for the information Matej,

Our stats seem to agree with yours regarding the bytes ratio, but there is a significant difference in the flows ratio.
Taking into account that we're still evaluating the flows' measurements, there is a slight possibility we might be doing something wrong.
Unless our users are opening too many (not http) connections.

--
Tassos

Matej Gregr wrote on 26/06/2013 15:48:
> Hello Tassos,
>   yes, we are doing the same measurements in our network and in the
> whole NREN network (CESNET) in the Czech Republic.
> You can access the statistics from our network at
> http://6lab.cz/live-statistics/ipv6-brno-univeristy-of-technology/ and
> for the whole NREN network at
> http://6lab.cz/live-statistics/ipv6-cesnet-nework/
>
> True is, that the statistics at 6lab.cz are not from dual stack only
> network, but from the whole networks. I have done measurements only from
> dual stack network (6000 users) with the following results.
>
> Around 80 % of users in the dual stack network are configured with IPv4
> and IPv6, the rest are only IPv4 (mainly Windows XP).
>
> The IPv4/IPv6 ratios of flows and bytes are following.
>
> IP4 flows IN ratio:  85.7 %        IP4 bytes IN ratio: 83 %
> IP6 flows IN ratio:  14.3 %        IP6 bytes IN ratio: 17 %
> IP4 flows OUT ratio: 85.7 %        IP4 bytes OUT ratio: 93 %
> IP6 flows OUT ratio: 14.3 %        IP6 bytes OUT ratio: 7%
>
> Greets,
>
> M.
>
>
>
>
> On 06/26/2013 11:28 AM, Tassos Chatzithomaoglou wrote:
>> I don't know if anyone is doing such measurements or have similar data, but i'm looking for the following:
>>
>> percent rate of number of IPv6 flows VS number of IPv6+IPv4 flows for Dual-Stack subscribers, how it evolves over time
>> percent rate of IPv6 download rate VS IPv6+IPv4 download rate for Dual-Stack subscribers, how it evolves over time
>>
>> I'm actually trying to come up with an algorithm that calculates the difference in resource (flows,bandwidth) usage of NAT444 vs IPv6+NAT44 subscribers.
>> I've made some calculations based on our own numbers (currently 5% and 17%), but it would be good to get input from others too.
>>
>> Thanks,
>> Tassos
>>
>