Suboptimal behavior from nsd

[Roy Arends]

On Thu, 15 Jan 2004, Miek Gieben wrote:

> [On 15 Jan, @12:01, Roy wrote in "Re: Suboptimal behavior from n ..."]
> >
> > > I don't think fixing this is trivial in the current
> > > NSD design (current: 1.2.X and 2.X.X).
> >
> > But is it going to be fixed ?
>
> short answer: no
> slight longer answer: no :-)

Ack.

> > > > from a single zone. This design is not spoof-proof.
> > >
> > > a well implemented cache should see that and not cache that information. It
> > > sounds a bit strange in my ears to talk about spoof-proofing NSD while
> > > NSD has no cache...
> >
> > ....
> >
> > Relying on a well implemented cache is a wee bit naive. It think both the
> > spoofer and the spoofee carry some responsibility for spoof-proofness.
>
> true,

Ack.

> but NSD added it because the glue was in file and on disk, so from nsd's
> viewpoint it as authoritative (and correct) as hell.

Whatever.

Lets do beer next week.

Roy