You are previewing Architecting High Availability Using WebSphere V6 on z/OS.
O'Reilly logo
Architecting High Availability Using WebSphere V6 on z/OS

Book Description

In this IBM Redbooks publication, we describe how to configure the various components of an e-business solution to exploit the availability and scalability benefits of zSeries and Parallel Sysplex using multiple LPARs running Linux for zSeries and z/OS.
This publication applies to WebSphere for z/OS V6, and is a continuation of the project evaluating high availability using WebSphere for z/OS V5 and V4. Presented in three parts (theory, systems setup, and availability tests), we cover workload balancing, the use of HTTP sessions, various architectures, and how to set up and test your infrastructures.
Considerations for configuring the systems and applications in an e-business environment are also examined, and we address such questions as:
- Should there be separate front-end systems for running the WebSphere Application Server?
- How should e-business components be configured to minimize or eliminate the impact of a system outage?
- What is the best way to communicate between the WebSphere systems and the back-end application systems?
- What is the impact on the end user if there is a failure in one of the front-end or back-end systems?

Table of Contents

  1. Notices
    1. Trademarks
  2. Preface
    1. The team that wrote this redbook
    2. Become a published author
    3. Comments welcome
  3. Part 1: Theory
    1. Chapter 1: Introduction
      1. Availability
        1. zSeries availability
        2. z/OS availability
        3. WebSphere Application Server for z/OS availability
        4. Linux for zSeries availability
        5. WebSphere Edge Components
        6. Firewall availability
      2. Scalability
        1. zSeries scalability
        2. z/OS scalability
        3. WebSphere Application Server for z/OS scalability
        4. Linux for zSeries scalability
      3. J2EE application characteristics
      4. Load balancing and affinity
        1. Our environment
        2. Load balancing with session persistence
        3. Load balancing with HTTP session affinity
    2. Chapter 2: Considerations for scalability
      1. Definitions
        1. Performance
        2. Scalability
        3. Throughput
        4. Load balancing and workload management
        5. Availability
        6. Recovery
        7. Maintainability
        8. Session management
        9. Security
      2. Increasing performance
        1. Performance definition
        2. Performance measurement
        3. Tuning performance
      3. Scaling your environment
        1. WebSphere scaling anatomy
        2. Scaling process
        3. Understanding your workload
        4. Scaling techniques
      4. Increasing throughput
        1. Caching
      5. Application programming patterns
        1. Applicable patterns
        2. Other considerations
    3. Chapter 3: Workload balancing
      1. Workload balancing
      2. Domain Name Server mapping solutions
        1. Round-robin DNS
        2. Connection optimization (DNS/WLM)
      3. Connection dispatching solutions
        1. Load Balancer
        2. Sysplex Distributor
        3. Sysplex Distributor and MultiNode Load Balancing (MNLB)
    4. Chapter 4: HTTP sessions
      1. Overview
      2. Maintaining session state
        1. Client tier
        2. Web tier - HTTP sessions
        3. Session affinity
        4. Session recovery and failover
        5. Database session persistence
        6. Memory-to-memory session replication
      3. Data flow in HTTP sessions with affinity
        1. Session persistence using an internal replication domain
        2. HTTP session failover tests
      4. HTTP sessions versus stateful session beans
    5. Chapter 5: Architectures for availability
      1. Things to consider
      2. Logical tiers
      3. Logical architectures
        1. Two-tier logical architecture
        2. Three-tier logical architecture
      4. Physical architectures
      5. z/OS base infrastructure
      6. Two-tier physical architecture
      7. Three-tier physical architectures
        1. Decouple Web/EJB and EIS layers
        2. Decouple Web and EJB layers
    6. Chapter 6: WebSphere platform messaging
      1. Overview
      2. Concepts and architecture
        1. Buses
        2. Bus members
        3. Messaging engines
        4. Data stores
        5. Destinations
        6. Mediations
        7. Foreign buses
      3. Availability
        1. Failover
        2. Messaging engine clustering
      4. Scalability
      5. Workload balancing
        1. Connecting to a service integration bus
        2. Push workload balancing
        3. Pull workload balancing
        4. WLM Classifier
      6. Core group policies
      7. Things to consider when writing applications
  4. Part 2: System setup
    1. Chapter 7: Setting up the infrastructure
      1. Infrastructure considerations
      2. The architecture we used
        1. The front-end systems
        2. The back-end systems
      3. Setting up the Network Dispatcher
        1. Linux for zSeries installation
        2. Configuration methods
        3. Load Balancer configuration
        4. High availability configuration
        5. Web server forwarding configuration
        6. z/Series TCP configuration (Layer 2 vrs Layer 3)
      4. Setting up the IBM HTTP Server on zLinux
        1. IBM HTTP Server configuration
        2. WebSphere Edge Components/Web server plug-in
        3. Configuring the WebSphere plug-in
        4. How the front-end traffic flows
    2. Chapter 8: Setting up the z/OS infrastructure
      1. Sysplex Distributor
        1. The implementation we used
      2. Setting up WebSphere
        1. Target infrastructure
        2. HFS setup for the SMP/E environment
        3. Setup of the WebSphere libraries
        4. WebSphere configuration
        5. IP name on the Location Service daemon definition
        6. Port definitions and assignments
        7. Define the first server
        8. Modifying the port settings
        9. New server RACF definitions
        10. Virtual host alias definitions
        11. Defining the cluster
        12. Web container definitions
        13. How these definitions fit together
      3. Other options to set up
        1. Setup for memory-to-memory internal replication
        2. Setup for session persistence with DB2
        3. Peer Restart and Recovery function
        4. Sysplex Failure Management (SFM)
      4. Maintenance strategy
        1. The HFS structure for upgrades
        2. Applying maintenance
        3. HFS versus zFS
    3. Chapter 9: Setting up the messaging infrastructure
      1. Set up the data stores
        1. Create the data store tables
        2. Create the JDBC data source for the messaging engines
      2. Configure the service integration bus (SIB)
        1. Create the bus
        2. Add the cluster to the bus
        3. Add additional messaging engines
      3. Configure preferred servers
        1. Set up core group policies
      4. Verify the configuration
  5. Part 3: Availability tests
    1. Chapter 10: Applications used to drive the workload
      1. Applications used to create the workload
        1. TraderDB2
        2. TraderJMS
      2. WebSphere Workload Simulator
        1. The workload
        2. Scripts for WebSphere Studio Workload Simulator
      3. Setting up TraderJMS
      4. Configuring the required resources
        1. Configuring service integration bus (SIB) resources
        2. Configure default messaging provider resources
        3. Configure JDBC resources
      5. Install the TraderJMS application
    2. Chapter 11: Planned infrastructure outages
      1. Installing new maintenance levels of WebSphere for z/OS
        1. Our installation’s example to roll the upgrade
        2. Implementing the upgrade on system SC54
        3. Rolling Upgrade process
    3. Chapter 12: Unplanned outages
      1. Summary of unplanned outage tests
      2. Edge server
      3. Proxy server
      4. HTTP server
      5. Single WebSphere
      6. WebSphere servant process
      7. Two WebSpheres
      8. Sysplex Distributor
      9. Deployment Manager with Peer Recovery
      10. WebSphere control region adjunct
      11. Summary of test results
      12. End-to-end monitoring
    4. Chapter 13: Application availability tests
      1. New Application Update features
        1. Finer grain application update
        2. Rollout Update
      2. Our test
        1. Rollout Update of the entire TraderDB ear file
        2. Single module update
        3. Single file update
    5. Related publications
      1. IBM Redbooks
      2. Other publications
      3. Online resources
      4. How to get IBM Redbooks
      5. Help from IBM
    6. Index
    7. Back cover