[194185] in North American Network Operators' Group
Re: Facebook more specific via Level3 ?
daemon@ATHENA.MIT.EDU (Mike Hammett)
Wed Mar 22 07:25:59 2017
X-Original-To: nanog@nanog.org
Date: Wed, 22 Mar 2017 06:25:49 -0500 (CDT)
From: Mike Hammett <nanog@ics-il.net>
Cc: nanog@nanog.org
In-Reply-To: <1490176932.2483603.919494848.3B91E17E@webmail.messagingengine.com>
Errors-To: nanog-bounces@nanog.org
Are your DNS resolvers on your network? No DNS forwarding?=20
-----=20
Mike Hammett=20
Intelligent Computing Solutions=20
Midwest Internet Exchange=20
The Brothers WISP=20
----- Original Message -----
From: "Radu-Adrian Feurdean" <nanog@radu-adrian.feurdean.net>=20
To: "J=C3=BCrgen Jaritsch" <juergen@jaritsch.at>, "Doug Porter" <dsp@fb.com=
>, nanog@nanog.org=20
Sent: Wednesday, March 22, 2017 5:02:12 AM=20
Subject: Re: Facebook more specific via Level3 ?=20
On Tue, Mar 21, 2017, at 20:38, J=C3=BCrgen Jaritsch wrote:=20
> I understand that FB is using some type of DNS geo-loadbalancing and othe=
r=20
> mechanism to redirect users to (possibly) nearer mirrors. The used DNS is=
=20
> directly requesting the root DNS and not any other public DNS (e.g. not=
=20
> 8.8.8.8). Running some requests within 3 minutes gives me the below=20
> results:=20
>=20
> www.facebook.com =3D> star-mini.c10r.facebook.com. =3D> 31.13.77.36=20
> www.facebook.com =3D> star-mini.c10r.facebook.com. =3D> 157.240.2.35=20
> www.facebook.com =3D> star-mini.c10r.facebook.com. =3D> 31.13.93.36=20
> www.facebook.com =3D> star-mini.c10r.facebook.com. =3D> 31.13.76.68=20
Hi, the load-balancing definitely doesn't choose the *nearest* mirror.=20
We are in France and unless we do dirty tricks, we *always* get directed=20
to US sites (as far as LA), with horrible performance. Everything since=20
end of December. As a consequence we let the dirty tricks in place=20
(query facebook.com and fbcdn.com on 8.8.8.8 instead of regular=20
recursive resolving) and we get directed to Frankfurt or Amsterdam=20
(never London or Paris).=20
