Browse Prior Art Database

RBridge Traffic Aggregation and Compression

IP.com Disclosure Number: IPCOM000173680D
Original Publication Date: 2008-Aug-20
Included in the Prior Art Database: 2008-Aug-20
Document File: 4 page(s) / 193K

Publishing Venue

Motorola

Related People

Donald E. Eastlake 3rd: INVENTOR

Abstract

RBridges form ideal concentration points within a network to apply compression and/or aggregation to traffic, improving efficiency. RBridges are a new technology being standardized in the IETF to bring the benefits of link state routing to layer-2 frame forwarding. A set of links and/or bridged LANs connected by RBridges appear to be a single logical link and broadcast domain to layer-3 devices such as routers; however, RBridges terminate bridge spanning trees and use a more sophisticated protocol with each other which encapsulates Rbridge-to-RBridge forwarded traffic. They provide optimum pair wise paths as well as Internet Protocol multicast and other optimizations and support multi-pathing.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 32% of the total text.

RBridge Traffic Aggregation and Compression

By Donald E. Eastlake 3rd

Motorola, Inc.

Broadband Network Research Laboratory, Motorola Laboratories

 

 

ABSTRACT

RBridges form ideal concentration points within a network to apply compression and/or aggregation to traffic, improving efficiency. RBridges are a new technology being standardized in the IETF to bring the benefits of link state routing to layer-2 frame forwarding. A set of links and/or bridged LANs connected by RBridges appear to be a single logical link and broadcast domain to layer-3 devices such as routers; however, RBridges terminate bridge spanning trees and use a more sophisticated protocol with each other which encapsulates Rbridge-to-RBridge forwarded traffic. They provide optimum pair wise paths as well as Internet Protocol multicast and other optimizations and support multi-pathing.

PROBLEM

Network bandwidth is limited by the overhead involved per unit of data (“frame”) transferred. This overhead can dominate network traffic for short frames but is reduced when multiple frames can be aggregated. Network bandwidth can also be increased in some cases by data compression but the gains with compression are larger with larger frames so the benefits of compression are multiplied by aggregation. If aggregation is performed, the aggregated frames must be destined for the same address and be disaggregated before delivery.

With conventional bridges, destinations are fine-grained individual MAC addresses, limiting opportunities for aggregation.

 

SOLUTION

RBridges form natural concentration points through which frames flow as shown in the figure above. Furthermore, the “ingress” RBridge where a frame is encapsulated and any “egress” RBridge or RBridges, which de-capsulate a frame, must already examine those frames and perform table look up based on MAC address, etc. As such these ingress and egress RBridges are ideal places at which to perform any or several of the following:

1.         aggregate small frames having the same RBridge destination into a larger frame,

2.         apply header compression to the multiple headers encapsulated for such aggregations,

3.         or to apply general data compression.

            In the absence of RBridges, this sort of aggregation by bridges would generally have to be based on MAC addresses so that multiple frames destined for the same MAC address could be aggregated to be split apart when finally delivered. But typically RBridges that have end devices attached will have many attached devices with different MAC addresses. (One RBridge interface could go to a classically bridged network with a vast number of devices.) RBridge to RBridge taffic, which is sent to a particular destination RBridge (or set of RBridges for multicast/broadcast), is much more likely to have easily...