[188542] in North American Network Operators' Group
Re: Some doubts on large scale BGP/AS design and black hole routing
daemon@ATHENA.MIT.EDU (Bill Woodcock)
Sun Apr 3 23:29:00 2016
X-Original-To: nanog@nanog.org
From: "Bill Woodcock" <woody@pch.net>
In-Reply-To: <BLU436-SMTP4563E5D7AB798A68A82915B0990@phx.gbl>
Date: Sun, 3 Apr 2016 20:28:50 -0700
To: "magicboiz@hotmail.com" <magicboiz@hotmail.com>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
Are respondents to suppose that the customer base and address space are even=
ly divided between the two cities, and that the ISP is too clueless to origi=
nate each /23 from the city that uses it, in iBGP?
=20
-Bill
> On Apr 3, 2016, at 15:04, "magicboiz@hotmail.com" <magicboiz@hotmail.com> w=
rote:
>=20
> Hi everybody!
>=20
>=20
> as part of laboratory work at the university, I'm working on a BGP design=
study, and I would like to post some questions regarding IP address space a=
llocation and its impact on BGP which are breaking my mind :)
>=20
> Let's suppose we have an ISP/AS with two POPs: PARIS and LONDON. These two=
POPs are connected with redundant leased lines. Each POP has a BGP router s=
peaking eBGP to different ISP providers/upstreams and also, each POP run its=
own OSPF area/ISIS area. Something like this:
>=20
>=20
> <INET> ---eBGP---<LONDON POP-ospf area1>=3D=3D=3Dredundant leased lines (o=
spf area0)=3D=3D=3D<PARIS POP- ospf area2>---eBGP---<INET>
>=20
> Now, this AS/ISP gets one /22 prefix from it RIR (RIPE in this case), and s=
tarts to announce it to its upstreams in PARIS and LONDON at the same time.
>=20
>=20
> My questions are:
>=20
> 1. What could happen in the case of total failure in the redundant leased l=
ines? Black hole routing between POPs?
>=20
> 2. What are the best design methods to avoid this scenario?
>=20
> 2.1: adding a third POP creating a triangle? What if a POP looses connec=
tion with the other two POPs at the same time? Another black hole?
>=20
> 2.2: requesting another prefix and allocating 1:1 prefix:POP, so in the s=
cenario each POP only would announce its prefix to the upstreams?
>=20
> 2.3: other?
>=20
>=20
>=20
> Thanks in advance!
> J.
>=20
