[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

quality of v6 connections [Re: unmanaged scope comments]



I'll revisit this due to a private comment.

Perhaps I should clarify this a bit.

Your take seems to be "how do we solve this problem given the network and 
v6 connectivity we have".

Mine is basically "we have a problem with network connectivity and v6 and
with the current state, turning v6 on for everything will produce
dissatisfied users".

With "network connectivity" I'm referring to the 6bone mess.

People usually just say, "enable IPv6" but it's more complex that.  As an 
user that actually uses IPv6 quite a bit, I'd be very hesitant to enable 
full use in a generic form.

I'm not sure whether trying to solve this is in our charter or not (or 
even if it was, would we be able to do anything), but that's one real 
problem.  6bone mess is hindering IPv6 deployment.

FWIW, I think 6bone addresses etc. are useful to get people started and 
acquainted with IPv6.  But tunnel-to-everyone, transit-to-everyone mess is 
not.  Those 6bone sites should be connected as leaf sites to proper 
connectivity providers.

On Thu, 19 Sep 2002, Pekka Savola wrote:
> On Fri, 20 Sep 2002, Jun-ichiro itojun Hagino wrote:
> > >==> connectivity includes one very important component IMO: the quality of
> > >network connections.  This is often ignored.  It's a high risk for a vendor
> > >to enable IPv6 by default, for example, if that'd result in lower quality
> > >connections to e.g. dual-stack web servers (a very real fact in 6bone), this
> > >should be taken into consideration at least in the short term.
> > 
> > 	i don't understand what you want to do about it.  will you ping/ping6
> > 	destination and pick shorter roundtrip?  when multiple A records
> > 	are returned for www.foo.com, do you probe them to get the optimal
> > 	performance?  
> 
> That's a difficult, general question to consider.  It's not easy.  I've 
> said one very real possibility to *not* introduce production services in 
> the same DNS names yet..
> 
> > i believe the "quality" factor does not matter, and
> > 	it cannot be measured from endpoint (at least within short time).
> 
> Connection to U.S. from Europe being 400 ms (and a few packets getting
> lost here and there) versus 150 ms is very measurable even without any
> tools e.g. in web browsing.
> 
> 

-- 
Pekka Savola                 "Tell me of difficulties surmounted,
Netcore Oy                   not those you stumble over and fall"
Systems. Networks. Security.  -- Robert Jordan: A Crown of Swords