Re: [rrg] procedural aggregation

Jakob Heitz <> Wed, 12 March 2014 03:35 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id 1E6BF1A039A for <>; Tue, 11 Mar 2014 20:35:47 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -0.002
X-Spam-Status: No, score=-0.002 tagged_above=-999 required=5 tests=[BAYES_20=-0.001, SPF_PASS=-0.001] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id 0Mt29GWptPov for <>; Tue, 11 Mar 2014 20:35:45 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 2AB961A0398 for <>; Tue, 11 Mar 2014 20:35:45 -0700 (PDT)
X-AuditID: c6180641-b7f2f8e000002cdc-77-531fd322f779
Received: from (Unknown_Domain []) by (Symantec Mail Security) with SMTP id 8E.DD.11484.223DF135; Wed, 12 Mar 2014 04:23:14 +0100 (CET)
Received: from ([]) by ([]) with mapi id 14.02.0387.000; Tue, 11 Mar 2014 23:35:37 -0400
From: Jakob Heitz <>
To: William Herrin <>, RRG <>
Thread-Topic: [rrg] procedural aggregation
Date: Wed, 12 Mar 2014 03:35:36 +0000
Message-ID: <>
References: <> <> <> <> <> <> <> <>
In-Reply-To: <>
Accept-Language: en-US
Content-Language: en-US
x-originating-ip: []
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFjrELMWRmVeSWpSXmKPExsUyuXSPn67SZflggyP9FhaHbh1ktFi04imL A5PHjpUVHpM3HmYLYIrisklJzcksSy3St0vgyvhy/Td7QRNvxbHFS5gaGFdydTFyckgImEhM 3LWZGcIWk7hwbz1bFyMXh5DAEUaJbaevQjnLGSWWr77GCFLFJqAj8e16F1iHiIC5xNcXm9lA bGEBTYnjv7+yQMS1JFbtOcQEYSdJ3Hv0D6yGRUBVYv2VE2A2r4CvxLf1CxkhFrxjlvi2+wDY UE6BQIl96/6CLWMEOun7qTVgg5gFxCVuPZnPBHGqgMSSPeehzhaVePn4HyuErSTx8fd8doh6 HYkFuz+xQdjaEssWvmaGWCwocXLmE5YJjKKzkIydhaRlFpKWWUhaFjCyrGLkKC1OLctNNzLc xAiMh2MSbI47GBd8sjzEKM3BoiTO++Wtc5CQQHpiSWp2ampBalF8UWlOavEhRiYOTqkGxkDP Wh0ZR8lEh43L4zjOLjcvnFr7RWaWV+kam52TOnm2LNTPCX7bHRCisvHy1DDLeOZpV6qWSqWH LL0TwTO3se5NPpOI0pHpfQc8ZM/Nn7VXstypqKhu6c3TkpIPljYKLTilfLTi8ccH+aI590rE RRfVMAbUsm2RFuX65FtTvbFwU6la2O/nSizFGYmGWsxFxYkAwBljwFUCAAA=
Subject: Re: [rrg] procedural aggregation
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: IRTF Routing Research Group <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Wed, 12 Mar 2014 03:35:47 -0000

No, he doesn't.
W draws a gige from A
How A connects to C is none of W's business.
If it's through B, then it is for A and B to negotiate a suitable peering agreement.
W is using B as backup and wants no traffic from it unless A goes down.
How do you think the traffic should flow in this case?


-----Original Message-----
From: rrg [] On Behalf Of William Herrin
Sent: Tuesday, March 11, 2014 7:59 PM
Subject: Re: [rrg] procedural aggregation

Hi Folks,

Here's another circumstance worth mentioning. This particular one is an undesirable case that exists in BGP today. It would be better if it didn't work. When I get to procedural aggregation, it'll be important to consider whether my notions make this scenario worse.

You have three networks: A, B and C. A peers with B, B peers with C.


You have a customer, W. W buys transit service from A (gige) and B (T1).

| /

W announces to B and announces plus to A.

Because W is a customer, B announces to C but as soon as the packet arrives at B from C, B sends it to A because of the more specific route it heard from A.

This allows W to draw a gige of traffic through B even though he's only paying for a T1.

Bill Herrin

William D. Herrin ................
3005 Crane Dr. ...................... Web: <> Falls Church, VA 22042-3004

rrg mailing list