[122267] in North American Network Operators' Group
Re: The Internet Revealed - A film about IXPs v2.0: now available
daemon@ATHENA.MIT.EDU (Michael Hallgren)
Wed Feb 10 17:07:32 2010
From: Michael Hallgren <m.hallgren@free.fr>
To: Nick Hilliard <nick@foobar.org>
In-Reply-To: <4B72D675.6000908@foobar.org>
Date: Wed, 10 Feb 2010 23:05:07 +0100
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
--=-JcRgb59q2d/yhQleLKl1
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Le mercredi 10 f=C3=A9vrier 2010 =C3=A0 15:53 +0000, Nick Hilliard a =C3=A9=
crit :
> On 10/02/2010 14:46, Mikael Abrahamsson wrote:
> > I guess we can agree to disagree then. I think it's highly biased
> > towards promoting IXPs,=20
>=20
> Uh, it was produced and paid for by IXPs for the intention of promoting
> IXPs. Why do you have an issue with this?
>=20
> > and it gives the impression that private peering
> > isn't settlement free and that it can't be used to do what an IXP does.
> > It just doesn't say so explicitly, but implies that it is so by the flo=
w
> > of how things are said and in what order. It sets private connects
> > against IXPs, and then describes all things an IXP can be used for, thu=
s
> > giving the impression that the PNI can't do this.
>=20
> Call me glib, but if you can get the association of PNI providers togethe=
r
> to create a movie about what PNIs are and how they work, I'd be ok if the=
y
> glossed over IXPs.
Good point.
>=20
> > But one factual error for instance, a TCP session (a picture being
> > transfrred) doesn't take multiple paths, that's just wrong to say so.
>=20
> ECMP? Per packet load balancing, even? Again, the point they were makin=
g
> is that the path from A to B is not particularly important to the data
> being transferred.
>=20
> Look, the creators of the movie had 5 minutes to explain something so tha=
t
> regular Janes and Joes would understand, rather than 1 hour to give a ner=
dy
> in-depth explanation of the nuts and bolts of IXPs. Personally, I think
> they did a rather good job.
>=20
So do I.
Cheers,
mh
> Nick
> (day job: contract IXP operations)
>=20
--=-JcRgb59q2d/yhQleLKl1
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Ceci est une partie de message
=?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iEYEABECAAYFAktzLY0ACgkQZNZ/rrgsqacGlwCgkvqLBUHjP0gfjIurN1EbL+t1
+qEAnjUFkIdqiUS2ZFX6MFW61wZiyFnH
=lg1N
-----END PGP SIGNATURE-----
--=-JcRgb59q2d/yhQleLKl1--
