Re: [DNSOP] [homenet] Fwd: WGLC on "redact" and "homenet-dot"

[Shane Kerr <[e-mail]>]

Bill,

At IETF 96 in Berlin, Warren gave a presentation discussing how Google
is using this in their recursive servers. Here's the link to the
recorded video for the whole dnsop session:

http://recs.conf.meetecho.com/Playout/watch.jsp?recording=IETF96_DNSOP&chapter=chapter_1

For me the most interesting slide is around minute 25 where we see
aggressive NSEC reducing the number of queries from Google servers to
the root by over 90%.

Cheers,

--
Shane

At 2016-12-17 12:17:17 -0800
william manning <[e-mail]> wrote:

> David, it would seem that fact-driven processes might serve the operational
> ecosystem better than SWAG, don't you agree?
> Warren, do you have, even antecdotal data on the impact of aggressive NSEC
> and traffic to the roots, that would inform this discussion (maybe).  At
> least it would give the root operators a heads up that they were, once
> again, being asked to backstop issues that should be handled elsewhere.
> 
> /Wm
> 
> On Sat, Dec 17, 2016 at 11:42 AM, David Conrad <[e-mail]> wrote:
> 
> > Bill,
> >
> > On Dec 17, 2016, at 11:36 AM, william manning <[e-mail]>
> > wrote:  
> > > Is there any public data to support the presumptions of excess capacity  
> > at the roots and the impact of NSEC aggressive use on the DNS?
> >
> > Warren provided some interesting anecdotes at the last IEPG, but I'm
> > unaware of any formal modeling.
> >  
> > > I know that in the previous century, punting on operational impact by  
> > guessing about outcomes was common.   I thought the IETF had moved away
> > from SWAG and was working toward a more disciplined and fact based process
> > for making changes.
> >
> > I make no comment on what the IETF has moved towards or away from.
> >
> > Regards,
> > -drc
> > (speaking only for myself)
> >
> >
> >
> >