A preliminary study of batch programming using WebSphere Extended deployment Compute grid

Introduction

Compute Grid is a feature of the IBM WebSphere Extended deployment V6.1 that provides the most complete Enterprise Java batch programming solution. Through Compute Grid, you can get:

Concise but also very powerful programming model based on traditional Java objects (plain old Java Object,pojo).

Simple packaging.

A simple deployment model.

Full-function Job Control language (Job Controls LANGUAGE,JCL).

A mature job scheduler.

A reliable execution environment.

Comprehensive workload management and management tools.

Although Compute Grid is designed to collaborate with other WebSphere Extended deployment features, it can be purchased and deployed independently. In a production environment, Compute Grid operates using the IBM WebSphere application Server Network Deployment (which uses a distributed, multiple-computer configuration), but Compute Grid also provides a unit test environment, You can run a stand-alone WebSphere application Server in it. Compute Grid also provides a development experience based on Eclipse and supports the use of IBM rational®application Developer as a full-featured development environment.

This article explains how to use Compute Grid for Java batch programming. However, you must first be fully aware of the batch job and the programming model that Compute Grid provides for building batch applications. After discussing these two topics, this article will guide you through the development of a simple batch application using the Compute Grid and the Batch emulator Test utility.

Analysis of batch processing job

From a more abstract perspective, a batch job is a declarative construct that instructs the execution of a sequence of one or more batch applications and specifies its inputs and outputs. batch jobs perform this set of tasks sequentially to complete specific business functions. A batch application is a program that is designed to be executed in the background in a way that is not iterative. Input and output are often used as logical constructs for batch application access, and are mapped to specific data resources by a batch job definition.

batch jobs typically deal with large amounts of input/output data (usually oriented to records) that are typically business-critical data, such as customer accounts, sales data, and so on. The scope of business processing tasks performed by batch operations is wide, including invoice generation, account optimization, business opportunity analysis, and so on. Batch tasks have been in the System Z (mainframe) environment for decades and continue to be used as central systems in many large and medium-sized enterprises today.

The basic content of a batch job consists of the components shown in Figure 1.

Figure 1. Analysis of batch processing job

The job definition describes the batch steps to be performed and the order in which they are run. Each step defines the specific batch application to invoke and its input and output data. Common sources and destinations for data include files, databases, transaction systems, message queues, and so on.