+ All Categories
Home > Documents > DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Date post: 13-Dec-2015
Category:
Upload: godfrey-manning
View: 213 times
Download: 0 times
Share this document with a friend
Popular Tags:
16
DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu
Transcript
Page 1: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

DYNAMIC LOAD BALANCINGON WEB-SERVER SYSTEMS

byValeria CardelliniMichele Colajanni

Philip S. Yu

Page 2: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Contents INTRODUCTION CLIENT-BASED APPROACH DNS-BASED APPROACH DISPATCHER-BASED APPROACH SERVER-BASED APPROACH COMPARING THE APPROACH CONCLUSIONS

Page 3: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

INTRODUCTION disproportionate increase in client requests to

popular Web sites solution

mirroring replicate information across a mirrored-server system not user-transparent

distributed Web-server system Information can be distributed among server nodes:

1st : content tree replication – LAN, WAN 2nd : information sharing – LAN

Successful load-balancing transparent users appear as a single host to outside world

four distributed Web-server architectures Client-based DNS-based , dispatcher-based , server-based

Page 4: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

CLIENT-BASED APPROCH Web Clients

Web client selects a node of the cluster and submits the request to the selected node Netscape Navigator – random select Limited practical applicability and is not scalable

Smart Clients Migrates server functionality to the client through a Java

applet Increase network traffic and network delay

Client-Side Proxies Web Cluster standpoint, proxy servers are similar to

clients not universally applicable

Page 5: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

DNS-BASED APPROACH

The cluster DNS translates URL to the IP address User transparent ex) www.yahoo.com www.cnn.com

Page 6: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

DNS-BASED APPROACH (con’t)

Drawbacks The DNS a limited control on the request reaching the Web clust

er Between the client and the web server DNS, many intermediate na

me servers can cache the logical name to IP address mapping to reduce network traffic and every web browser typically caches some address resolution

scheduling algorithm that the cluster DNS uses to balance the Web-server node’s load constant TTL algorims

System-stateless Server-state-based Client-state-based

adaptive TTL algorims DNS-Based Architecture Comparison

Page 7: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

DISPATCHER-BASED APPROACH Centralize request scheduling and completely control clie

nt-request routing Request routing among server is transparent-unlike DNS

-based DNS deals address at the URL level, the dispatcher has a single, virtua

l IP address(IP-SVA) Dispatcher uniquely identifies each server in the system

through a private address Dispatcher typically use simple algorithms to select the

Web server difference by routing mechanism

Packet Single-Rewriting Packet Double-Rewriting Pachet Forwarding (network dispatcher, One-ip address) HTTP redirection

Page 8: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Packet Single-Rewriting

TCP router acts as an IP address dispatcher High System availability

Page 9: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Packet Double-Rewriting

Two solution using this approach Magicrouter

round-robin, random, incremental load algrithm Cisco System’s Local Director

least number of active connections

Page 10: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Packet Forwarding

Network Dispatcher Dispatcher forward packets to the selected server using its

physical address without IP modification Level 1 : single-rewriting mechanisim Level 2 : LAN network dispatcher

Page 11: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

Packet Forwarding (con’t)

ONE-IP address uses the ifconfig alias option to configure a Web-serv

er system with multiple machines implemented with two techniques

Routing-based dispatching Broadcast-based dispatching

Page 12: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

HTTP Redirection

among the Web-server nodes through the HTTP’s redirection mechanism

no IP address modification two techniques

Server-state-based dispatching used by Distributed Server Groups architecture

Location-based dispatching used by Cisco System’s DistributedDirector appliance

Dispatcher-Based Architecture Comparison

Page 13: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

SERVER-BASED APPROACH Use two level dispatching mechanism

Integrating the DNS based approach with redirection techniques executed by Web server

Solves most DNS scheduling problem Two Solution

HTTP redirection

Packet redirection

Page 14: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

COMPARING THE APPROACHES Approach Trade-off Summary

Page 15: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

COMPARING THE APPROACHES Performance

Evaluation

Page 16: DYNAMIC LOAD BALANCING ON WEB-SERVER SYSTEMS by Valeria Cardellini Michele Colajanni Philip S. Yu.

CONCLUSIONS network bandwidth can constrain load-

balancing performance LAN-distributed Web-server cluster are

thus a limited solution to increased client requests


Recommended