Re: Comcast enables 6to4 relays

[Joel Jaeggli <joelja () bogus com>]

On 8/29/10 6:25 AM, John Jason Brzozowski wrote:
> Franck,
>
> As you know 6to4 is enabled by default in many cases and is used perhaps
> more than folks realize.  Because of this and other observations we decided
> to deploy our own relays.

Right prior to this the nearest 6to4 relay router from the vantage-point
of comcast customers was at AMSIX. It's a given that you're going to
have path asymmetry, in this case however it was frequently worse in one
direction than in the other.

This ought greatly improve the performance of existing devices located
at comcast's customers.

> This does not alter our plans for our native dual stack trials, in fact, I
> hope to have more news on this front soon.
>
> It is true that 6to4 has challenges, some of these may be related to how
> 6to4 relays have been deployed.  Others may be related to the protocol
> itself.  Either way, by deploying our own we observed an improvement, we
> hope others have as well.
>
> John
>
> On 8/28/10 6:06 PM, "Franck Martin" <franck () genius com> wrote:
>
> > These are good news.
> >
> > However, if Comcast provides native IPv6 to their customers, then the IPv6
> > native customers don't need these 6to4 relays?
> >
> > Airport Extreme, Linksys and other user equipment, enable IPv6 by doing 6to4
> > tunnels, so what this press release says, is that there are many users who are
> > already on IPv6 via Comcast network but not native? Providing relays close to
> > them, is a good transition move. Alternatively, the measurement of this 6to4
> > bandwidth on IPv4 may give you an idea of the demand for IPv6 from your
> > customers? May be you detected a non null number here?
> >
> > I'm just trying to understand more IPv6 by the examples.
> >
> > I'm personally using 6to4 at home, and experiencing some MTU issues, which
> > seems related to some PTB packets suppressed on the way between some end
> > points, and that can depend on which 6to4 relay I'm using. Still trying to
> > debug this (I'm not too fanatic about it, but work on it when I have a bit of
> > time). I thought I would mention that.
> > The WAND people have done some good studies:
> > http://www.ripe.net/ripe/meetings/ripe-60/presentations/Stasiewicz-Measurement
> > s_of_IPv6_Path_MTU_Discovery_Behaviour.pdf
> >
> > At the office, I have a more classical tunnel with he.net and do not have any
> > issue there.
> >
> > ----- Original Message -----
> > From: "John Jason Brzozowski" <john_brzozowski () cable comcast com>
> > To: "NANOG" <nanog () nanog org>
> > Sent: Sunday, 29 August, 2010 5:49:30 AM
> > Subject: Comcast enables 6to4 relays
> >
> > FYI - thought this would be of interest to some of you, there will be more
> > news on this front shortly.
> >
> > http://www.comcast6.net/
> >
> > 6to4 Relays Activated
> > Tuesday, August 17, 2010
> >
> > As we started our IPv6 trials, we began to observe an increase in 6to4 relay
> > traffic. 6to4 is a transition mechanism built into some operating systems
> > and home gateways. While it is not a transition technology that Comcast
> > planned to invest in due to limitations related to performance, we did
> > observe poor performance when 6to4 was used by our customers. In many cases,
> > these customers were not even aware that 6to4 was enabled by default or that
> > their device or operating system was attempting to use 6to4 to communicate
> > with IPv6 resources on the Internet.
>
> =========================================
> John Jason Brzozowski
> Comcast Cable
> e) mailto:john_brzozowski () cable comcast com
> o) 609-377-6594
> m) 484-962-0060
> w) http://www.comcast6.net
> =========================================