On the network impact of dynamic server selection

Abstract Widespread replication of information can ameliorate the problem of server overloading but raises the allied question of server selection. Clients may be assigned to a replica in a static manner or they may choose among replicas based on client-initiated measurements. The latter technique, called dynamic server selection (DSS), can provide significantly improved response time to users when compared with static server assignment policies (for example, based on network distance in hops). In the first part of this paper we demonstrate the idea of DSS using experiments performed in the Internet. We compare a range of policies for DSS and show that obtaining additional information about servers and paths in the Internet before choosing a server improves response time significantly. The best policy we examine adopts a strategy of never adding more than 1% additional traffic to the network, and is still able to provide nearly all the benefits of the most expensive policies. While these results suggest that DSS is beneficial from the network user's standpoint, the system-wide effects of DSS schemes should also be closely examined. In the second part of this paper we use large-scale simulation to study the system-wide network impact of dynamic server selection. We use a simulated network of over 100 hosts that allows local-area effects to be distinguished from wide-area effects within traffic patterns. In this environment we compare DSS with static server selection schemes and confirm that client benefits remain even when many use DSS simultaneously. Importantly, we also show that DSS confers system-wide benefits from the network standpoint, as compared to static server selection. First, overall data traffic volume in the network is reduced, since DSS tends to diminish network congestion. Second, traffic distribution improves – traffic is shifted from the backbone to regional and local networks.

[1]  David L. Peterson,et al.  Power Laws In Large Shop DASD I/O Activity , 1995, Int. CMG Conference.

[2]  Mark Crovella,et al.  Measuring Bottleneck Link Speed in Packet-Switched Networks , 1996, Perform. Evaluation.

[3]  Jean-Chrysostome Bolot,et al.  End-to-end packet delay and loss behavior in the internet , 1993, SIGCOMM '93.

[4]  Raj Jain,et al.  The Art of Computer Systems Performance Analysis : Tech-niques for Experimental Design , 1991 .

[5]  Srinivasan Keshav.  A control-theoretic approach to flow control , 1991, SIGCOMM 1991.

[6]  Van Jacobson,et al.  A tool to infer characteristics of internet paths , 1997 .

[7]  Azer Bestavros,et al.  Application-level document caching in the Internet , 1995, Second International Workshop on Services in Distributed and Networked Environments.

[8]  Martin F. Arlitt,et al.  Web server workload characterization: the search for invariants , 1996, SIGMETRICS '96.

[9]  Azer Bestavros,et al.  Demand-based document dissemination to reduce traffic and balance load in distributed information systems , 1995, Proceedings.Seventh IEEE Symposium on Parallel and Distributed Processing.

[10]  Michael F. Schwartz,et al.  Locating nearby copies of replicated Internet servers , 1995, SIGCOMM '95.

[11]  Vern Paxson,et al.  Measurements and analysis of end-to-end Internet dynamics , 1997 .

[12]  Mark Crovella,et al.  Server selection using dynamic path characterization in wide-area networks , 1997, Proceedings of INFOCOM '97.

[13]  David L. Peterson.  Data Center I/O Patterns And Power Laws , 1996, Int. CMG Conference.

[14]  Ellen W. Zegura,et al.  How to model an internetwork , 1996, Proceedings of IEEE INFOCOM '96. Conference on Computer Communications.

[15]  Lester Lipsky,et al.  The Importance of Power-tail Distributions for Telecommunication Traffic Models , 1995 .

[16]  Ellen W. Zegura,et al.  A novel server selection technique for improving the response time of a replicated service , 1998, Proceedings. IEEE INFOCOM '98, the Conference on Computer Communications. Seventeenth Annual Joint Conference of the IEEE Computer and Communications Societies. Gateway to the 21st Century (Cat. No.98.

[17]  Peter B. Danzig,et al.  A Hierarchical Internet Object Cache , 1996, USENIX ATC.

[18]  Van Jacobson,et al.  Congestion avoidance and control , 1988, SIGCOMM '88.

[19]  Azer Bestavros,et al.  Self-similarity in World Wide Web traffic: evidence and possible causes , 1997, TNET.

[20]  Walter Willinger,et al.  On the self-similar nature of Ethernet traffic , 1993, SIGCOMM '93.

[21]  Margo Seltzer,et al.  VINO: The 1994 Fall Harvest , 1994 .

[22]  Jean-Chrysostome Bolot,et al.  Characterizing End-to-End Packet Delay and Loss in the Internet , 1993, J. High Speed Networks.

[23]  Shuang Deng,et al.  Empirical model of WWW document arrivals at access link , 1996, Proceedings of ICC/SUPERCOMM '96 - International Conference on Communications.

[24]  Ellen W. Zegura,et al.  Application-layer anycasting , 1997, Proceedings of INFOCOM '97.

[25]  Margo I. Seltzer,et al.  The case for geographical push-caching , 1995, Proceedings 5th Workshop on Hot Topics in Operating Systems (HotOS-V).

[26]  Raj Jain,et al.  The art of computer systems performance analysis - techniques for experimental design, measurement, simulation, and modeling , 1991, Wiley professional computing.

[27]  Peter Scheuermann,et al.  Selection algorithms for replicated Web servers , 1998, PERV.

[28]  WillingerWalter,et al.  On the self-similar nature of Ethernet traffic , 1993 .