IBM WebSphere V5.1 performance, scalability, and high availability / [Birgit Roehm ... et al.].

Format:

Book

Contributor:

Roehm, Birgit.

International Business Machines Corporation. International Technical Support Organization.

Series:

WebSphere handbook series.

IBM redbooks.

WebSphere handbook series

IBM redbooks

Language:

English

Subjects (All):

Web servers--Computer programs.

Web servers.

Web site development.

WebSphere.

Physical Description:

xxvi, 980 p. : ill.

Edition:

2nd ed.

Place of Publication:

White Plains, NY : IBM, International Technical Support Organization, c2004.

Language Note:

English

Contents:

Front cover

Contents

Notices

Trademarks

Preface

The team that wrote this redbook

Become a published author

Comments welcome

Summary of changes

June 2004, Second Edition

Part 1 Getting started

Chapter 1. Overview and key concepts

1.1 Objectives

1.1.1 Scalability

1.1.2 Workload management

1.1.3 Availability

1.1.4 Maintainability

1.1.5 Session state

1.1.6 Performance impacts of WebSphere Application Server security

1.2 WebSphere Application Server architecture

1.2.1 WebSphere Application Server Network Deployment components

1.2.2 Web clients

1.2.3 Java clients

1.3 Workload management

1.3.1 Web server workload management

1.3.2 Plug-in workload management

1.3.3 Workload management using WebSphere clustering

1.3.4 Enterprise Java Services workload management

1.4 Managing session state among servers

1.4.1 HTTP sessions and the session management facility

1.4.2 EJB sessions or transactions

1.4.3 Server affinity

1.5 Performance improvements over previous versions

1.6 The structure of this redbook

Chapter 2. Design for scalability

2.1 Scaling your infrastructure

2.2 Understanding the application environment

2.3 Categorizing your workload

2.3.1 Workload patterns and Web site classifications

2.3.2 Workload characteristics

2.4 Determining the most affected components

2.5 Selecting the scaling techniques to apply

2.5.1 Using a faster machine

2.5.2 Creating a cluster of machines

2.5.3 Using appliance servers

2.5.4 Segmenting the workload

2.5.5 Batch requests

2.5.6 Aggregating user data

2.5.7 Managing connections

2.5.8 Caching

2.6 Applying the technique(s)

2.7 Re-evaluating

Chapter 3. Introduction to topologies

3.1 J2EE tiers model

3.2 Topology selection criteria

3.3 Strategies for scalability.

3.4 Single machine topology

3.5 Separating the Web server

3.6 Separating the database server

3.7 Vertical scaling

3.8 Horizontal scaling with clusters

3.8.1 Horizontal scaling with IP sprayer

3.9 One WebSphere administrative cell versus many

3.10 Multiple clusters on one node versus one cluster per node

3.11 The sample topology

3.12 Topologies and high availability

3.12.1 Using WebSphere Load Balancer custom advisor

3.13 Closing thoughts on topologies

3.14 Topology selection summary

Part 2 Distributing the workload

Chapter 4. Web server load balancing

4.1 Introduction

4.1.1 Scalability

4.1.2 Availability

4.1.3 Performance

4.2 IBM WebSphere Edge Components

4.3 Load Balancer overview

4.3.1 Load Balancer topologies

4.3.2 Installation and configuration

4.3.3 Setting up the cluster machines

4.3.4 Configuring a Web server cluster

4.3.5 Testing the configuration

4.4 Advisors

4.4.1 Custom advisors

4.4.2 Using WebSphere Application Server sample custom advisor

4.5 Server affinity

4.5.1 "Stickyness" to source IP address

4.5.2 Passive cookie affinity

4.5.3 Active cookie affinity

4.5.4 URI

4.5.5 SSL session ID

4.6 Caching Proxy

4.6.1 Forward proxy

4.6.2 Reverse proxy (IP forwarding)

4.6.3 Load Balancing

4.6.4 Dynamic caching

Chapter 5. Plug-in workload management and failover

5.1 Introduction

5.2 The embedded HTTP transport

5.3 Setting up the Web containers

5.3.1 Virtual hosts

5.3.2 Transports

5.3.3 Creating clusters and cluster members

5.4 WebSphere plug-in workload management

5.4.1 Processing requests

5.4.2 The plug-in configuration file

5.4.3 Generation of the plug-in configuration file

5.4.4 Plug-in workload management and failover policies

5.5 Session management

5.5.1 Session affinity.

5.5.2 Session failover inside the plug-in

5.5.3 Session identifiers

5.5.4 Session persistence and failover

5.6 Troubleshooting the Web server plug-in

5.6.1 Logging

5.6.2 Trace

5.7 Web server plug-in behavior and failover

5.7.1 Normal operation

5.7.2 Failover operation

5.7.3 Tuning failover

Chapter 6. EJB workload management

6.1 Enabling EJB workload management

6.2 EJB types and workload management

6.2.1 Stateless session beans

6.2.2 Stateful session beans

6.2.3 Entity beans

6.3 Naming and name spaces

6.3.1 Looking up an EJB home with JNDI examples

6.4 How EJBs participate in workload management

6.4.1 Initial request

6.4.2 Subsequent requests

6.4.3 Cluster run state changes

6.5 EJB server selection policy

6.5.1 Server weighted round robin routing configuration

6.5.2 Prefer local configuration

6.6 EJB workload management behavior

6.6.1 WLM behaviors using server weighted round robin

6.6.2 Prefer local

6.6.3 Process affinity

6.6.4 Transaction affinity

6.7 EJB workload management failover

6.7.1 Exceptions triggering automatic failover

6.7.2 Exceptions thrown by WLM to the application

6.8 Backup Cluster support

Part 3 Implementing the solution

Chapter 7. Implementing the sample topology

7.1 Overview

7.1.1 Software products

7.1.2 The sample topology

7.1.3 Applications used in our sample topology

7.2 Installation summary

7.3 Configuring Caching Proxy

7.3.1 Starting Caching Proxy

7.3.2 Set up Caching Proxy

7.3.3 Set administrator user ID and password

7.3.4 Restart Caching Proxy

7.4 Configuring Load Balancer

7.4.1 Configuring the Web servers for Load Balancer

7.4.2 Starting Load Balancer and administration GUI

7.4.3 Connecting to the Dispatcher host

7.4.4 Adding the Web server cluster.

7.4.5 Adding a port to the cluster

7.4.6 Adding the Web servers to the cluster

7.4.7 Start the Network Dispatcher manager

7.4.8 Checking what you have done until now

7.5 Configuring WebSphere clusters

7.5.1 Introduction

7.5.2 Creating the Web container cluster

7.5.3 Creating the EJB cluster

7.5.4 Configure persistent session management

7.6 Installing and configuring BeenThere

7.6.1 BeenThere installation summary

7.6.2 Install BeenThere

7.6.3 Regenerate Web server plug-in

7.6.4 Restart servers

7.6.5 Verifying BeenThere

7.7 Installing and configuring Trade3.1

7.7.1 Trade3.1 installation summary

7.7.2 Download the Trade3.1 package

7.7.3 Set up and configure Trade3DB database

7.7.4 Create JDBC and JMS resources

7.7.5 SOAPify Trade3.ear

7.7.6 Install Trade3.1 from the WebSphere Administrative Console

7.7.7 Regenerate Web server plug-in

7.7.8 Restart servers

7.7.9 Install Trade3.1 using the installation script

7.7.10 Working with Trade3.1

7.7.11 Verify failover with Trade3.1

7.7.12 Volume testing Trade3.1

Part 4 High availability solutions

Chapter 8. High availability concepts

8.1 Process availability and data availability

8.2 Clustering for high availability

8.3 Availability definition

8.3.1 Levels of availability

8.3.2 Availability matrix

8.3.3 Causes of downtime

8.3.4 Possible single points of failure in the WebSphere system

8.3.5 Levels of WebSphere system availability

8.3.6 Planning and evaluating your WebSphere HA solutions

8.4 Failover terms and mechanisms

Chapter 9. WebSphere Application Server failover and recovery

9.1 Overview

9.2 Web container clustering and failover

9.2.1 Web container failures and failover

9.2.2 Web server plug-in failover performance tuning

9.2.3 Network failures.

9.2.4 Stream and overloading failover

9.3 HTTP session failover

9.3.1 Session affinity and failover

9.3.2 Session update methods and failover session data loss

9.3.3 Session persistence and failover

9.4 EJB container failover

9.4.1 EJB client redundancy and bootstrap failover support

9.4.2 EJB container redundancy and EJB WLM failover support

9.4.3 EJB WLM routing

9.4.4 LSD failover

9.4.5 EJB container failover behavior and tuning

9.4.6 Fault isolation and data integrity

9.4.7 EJB caching and failover

9.4.8 EJB types and failover

9.4.9 Conditions of WLM failover

9.4.10 Resource redundancy (EJB database, JMS resource, LDAP)

9.5 Enhancing WebSphere HA using clustering software

9.5.1 Failover unit

9.5.2 Configuration and setup for HACMP

9.5.3 Failover process

9.5.4 Advantages

Chapter 10. Deployment Manager and Node Agent high availability

10.1 Introduction

10.2 Node Agent failures

10.2.1 Application servers

10.2.2 Deployment Manager

10.2.3 Location Service Daemon

10.2.4 Naming server

10.2.5 Security server

10.2.6 Application clients

10.2.7 Synchronization Service and File Transfer Service

10.2.8 RAS service, PMI and monitoring

10.2.9 Administrative clients

10.3 Enhancing Node Agent high availability

10.3.1 Add Node Agent as OS daemon

10.3.2 Enhancing Node Agent HA using clustering software

10.4 Deployment Manager failures

10.4.1 Configuration management

10.4.2 Node Agent

10.4.3 Application server

10.4.4 Naming server

10.4.5 Security server

10.4.6 WLM runtime service

10.4.7 Application clients

10.4.8 Synchronization Service and File Transfer Service

10.4.9 RAS Service and PMI monitoring

10.4.10 Administrative clients

10.5 Enhancing Deployment Manager high availability.

10.5.1 Add Deployment Manager to OS daemon service.

Notes:

"June 2004."

"SG24-6198-01."

Includes bibliographical references (p. 935-945) and index.

OCLC:

932363773