[187277] in North American Network Operators' Group
Re: The IPv6 Travesty that is Cogent's refusal to peer Hurricane
daemon@ATHENA.MIT.EDU (Mark Tinka)
Tue Jan 26 00:47:48 2016
X-Original-To: nanog@nanog.org
To: Brandon Butterworth <brandon@rd.bbc.co.uk>, jmaimon@ttec.com,
nick@foobar.org
From: Mark Tinka <mark.tinka@seacom.mu>
Date: Tue, 26 Jan 2016 07:47:00 +0200
In-Reply-To: <201601252228.WAA13974@sunf10.rd.bbc.co.uk>
Cc: nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
On 26/Jan/16 00:28, Brandon Butterworth wrote:
> Doesn't matter, if traffic is blackholed at an ix then it
> won't be failing over to another one. Same effect
Route servers do not participating in the forwarding plane. If they
fail, you lose routes from that exchange point which show up elsewhere.
If peers are originating routes at exchange points and lose their
backhauls, that's another set of problems your NOC can fix.
If the exchange point switch runs out of ideas, that's another set of
problems your NOC can fix.
> The general case doesn't care about your network, it assumes you'd
> engineer that appropriately for the criticality and do something
> different/better if you need to.
Big assumption to make.
Mark.
