09_5 - Scalability of Reliable Group Communication Using...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Scalability of Reliable Group Communication Using Overlays Franc ¸ois Baccelli Augustin Chaintreau Zhen Liu Anton Riabov †† Sambit Sahu INRIA & ENS 45, rue d’Ulm 75005 Paris FRANCE [email protected], [email protected] IBM T.J. Watson Research Center P.O. Box 704, Yorktown Heights NY 10598, USA [email protected], [email protected] †† Columbia University, IEOR Dept 540 W 120 st, New York NY 10027, USA [email protected] Abstract — This study provides some new insights into the scalability of reliable group communication mechanisms using overlays. These mechanisms use individual TCP connections for packet transfers between end-systems. End-systems store incoming packets and forward them to downstream nodes using different unicast TCP connections. In this paper we assume that buffers in end-systems are large enough for the transfers. It is shown that the throughput of the reliable overlay group communication scales in the sense that for all multicast tree sizes and topologies, the group throughput is strictly positive under natural conditions. This is in contrast with the IP supported multicast paradigm where reliable protocols have vanishing throughput when the group size tends to inFnity. The scalability of packet delay and buffer occupancy is then investigated. In the absence of additional control, the occupancy of the buffer and the latency in the end-systems explodes with time. It is then shown that proactive rate throttle mechanism implemented at the source leads to Fnite packet latency and buffer occupancy in any end-system of the network provided certain moment conditions are satisFed by cross trafFc in the routers. I. INTRODUCTION Reliable group communication has remained an important research problem for the last decade. Signi±cant effort has been spent on the design and evaluation of reliable multicast transport protocols, see for example [11], [7], [18] and the ref- erences therein. However, such IP supported reliable multicast schemes have been facing two major obstacles. First, there is no wide spread deployment of IP multicast in the Internet. Second, it has been shown in various studies [24], [9] that group throughput vanishes when the group size increases, thus suffering from scalability issues. Recently an alternative approach that uses overlays of end- systems has been proposed to support group communications. In this approach, end-systems form an overlay by establishing point-to-point connections in between end-systems, where each node forwards data to downstream nodes in a store-and- forward way. The multicast distribution tree is formed at the end-system level. Such a paradigm is referred to as end-system multicast, or application-level multicast, or simply multicast using overlays. Various studies have been conducted with the primary focus on the protocol development for ef±cient overlay tree construction and maintenance, such as Narada [10], Yoid [12], ALMI [23], Host Multicast [28], NICE [6], Delauney graph [19], and [26], [25].
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page1 / 12

09_5 - Scalability of Reliable Group Communication Using...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online