[126869] in North American Network Operators' Group
Re: Strange practices?
daemon@ATHENA.MIT.EDU (sjk)
Tue Jun 8 00:15:25 2010
Date: Mon, 07 Jun 2010 23:14:56 -0500
From: sjk <sjk@sleepycatz.com>
To: Bill Fehring <lists@billfehring.com>
In-Reply-To: <AANLkTikEhZyvFxJuu-ahFhcmFRCUr1z9YMRm9r6x03nU@mail.gmail.com>
Cc: nanog@nanog.org, "Murphy, Jay, DOH" <Jay.Murphy@state.nm.us>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Bill Fehring wrote:
> On Mon, Jun 7, 2010 at 14:59, Murphy, Jay, DOH <Jay.Murphy@state.nm.us> wrote:
>> "So if the enterprise loses connectivity to one of these two providers, does the provider without working connectivity to the enterprise have mechanism in place to cease originating the address space?"
>> Yes, BGP updates.
>
> Um, it wasn't a trick question Jay, and as others have stated, since
> the providers are statically routing this address space to their
> common customer, this would require a coordinated effort to manually
> (or preferably automatically) shutdown the advertisement should
> connectivity be lost to the customer. There are a number of ways that
> could be achieved, but it's obviously important that it is.
>
> -Bill
>
Not necessarily: the way that I have seen this implemented the upstreams
rely upon the static -- or sometimes connected -- route being pulled
from the route table if the interface goes down. Once pulled from the
table the it drops out of IGP and then from the eBGP announcement. It is
-- without a doubt -- a crappy solution as it doesn't deal with things
like looped circuits, bad encapsulations, etc....
