[171431] in North American Network Operators' Group
Re: We hit half-million: The Cidr Report
daemon@ATHENA.MIT.EDU (Patrick W. Gilmore)
Tue Apr 29 13:40:26 2014
X-Original-To: nanog@nanog.org
From: "Patrick W. Gilmore" <patrick@ianai.net>
In-Reply-To: <9BD28641-5B59-4C88-BD76-97F4E3BFA570@apnic.net>
Date: Tue, 29 Apr 2014 12:22:36 -0400
To: NANOG list <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
> The remainder of the prefixes (45%) shares the same origin AS and the =
same path.
> The could be TE prefixes, but as they are identical to their covering
> aggregate its hard to appreciate exactly what the engineering intent =
may be. I could
> make a wild guess and call these 45% of more specifics to be an act of =
senseless routing
> vandalism. ( :-) ) This number has been steady as a % for the past =
three years.
This could easily be TE, and a type of TE which would be trivially =
fixed.
Let's take a simple example of a network with a /22 and 4 POPs. They =
have the same transit provider(s) at all 4 POPs and a small backbone to =
connect them. Each POP gets a /24.
A not-ridiculous way to force their transit provider to carry bits =
instead of clogging their backbone while still ensuring redundancy would =
be to announce the /22 at all four POPs and the individual /24 at each =
individual POP. This creates four /24s and a covering /22 with exactly =
the same path, but still has "use" as TE.
Of course, it would be trivial for the network to clean up their act by =
attacking no-export to the /24s. But some people either do not know it =
exists, know how it works, or know BGP well enough to understand it =
would not harm them. Or maybe they are just lazy: "What's 3 extra =
prefixes in half a million?"
The answer to the last question is, frankly, nothing. But 3 prefixes for =
30K ASNs is an ass-ton. (That's a technical term meaning "lots & lots".)
This is a good time for a marketing effort. Let's see if we can get the =
table back under 500K. Everyone check your announcements. Are you =
announcing more specifics and a covering aggregate with the same path? =
Can you delete the more specific? Can you add no-export or another =
community to keep the more specifics from the global table?
If you are unsure, ask. I think it would be rather awesome if we saw a =
quick reversal in table growth and went back under 500K, even if it was =
short lived. ESPECIALLY if we can do it before we hit 512K prefixes. =
Would prove the community still cares about, well, the community, not =
just their own network. Because on the Internet, "your network" is part =
of the "community", and things that harm the latter do harm the former, =
even if it is difficult for you to see sometimes.
Who will be the first to pull back a few prefixes?
--=20
TTFN,
patrick
On Apr 29, 2014, at 03:31 , Geoff Huston <gih@apnic.net> wrote:
>=20
> On 29 Apr 2014, at 12:39 pm, Valdis.Kletnieks@vt.edu wrote:
>=20
>> On Mon, 28 Apr 2014 21:59:43 -0400, "Patrick W. Gilmore" said:
>>>> On Apr 28, 2014, at 19:41, Chris Boyd <cboyd@gizmopartners.com> =
wrote:
>>>> I'm in the middle of a physical move. I promise I'll take the 3 =
deagg'd
>>>> /24s out as soon as I can.
>>> Do not laugh. If everyone who had 3 de-agg'ed prefixes fixed it, the =
table
>>> would drop precipitously. We all have to do our part.
>>=20
>> Do we have a handle on what percent of the de-aggrs are legitimate =
attempts
>> at TE, and what percent are just whoopsies that should be =
re-aggregated?
>>=20
>=20
> I made a shot at such a number in a presentation to NANOG in Feb this =
year
> (http://www.potaroo.net/presentations/2014-02-09-bgp2013.pdf)
>=20
>=20
> If you assume that Traffic Engineering more specifics share a common =
origin AS with the
> covering aggregate, then around 26% of more specifics are TE =
advertisements. This=20
> number (as a percentage) has gwon by 5% over the past three years
>=20
>=20
> If you assume that Hole Punching more specifics are more specifics =
that use a different
> origin AS, then these account for 30% of the more specifics in today's =
routing table.
> This number has fallen by 5% over the past three years.
>=20
> The remainder of the prefixes (45%) shares the same origin AS and the =
same path.
> The could be TE prefixes, but as they are identical to their covering
> aggregate its hard to appreciate exactly what the engineering intent =
may be. I could
> make a wild guess and call these 45% of more specifics to be an act of =
senseless routing
> vandalism. ( :-) ) This number has been steady as a % for the past =
three years.
>=20
> Interestingly, it's the hole punching more specifics that are less =
stable, and the
> senseless routing vandalism more specifics that are more stable than =
the average.
>=20
> thanks,
> Geoff
