Spring Batch with Quartz scheduler

Introduction

Spring Batch is one of the most widely used frameworks for implementing an offline batch process. The programming model of the Spring Batch framework has been leveraged as a Java Platform Standard – JSR 352.

The scheduling of a Spring Batch can be done using any of the schedulers available in the market – Quartz, Control-M etc. Often for a basic scheduling, we can set up a crontab.

Below are the steps required for integrating Quartz scheduler with Spring Batch.

Configuration

We will be adding 2 new dependencies in pom.xml i.e. spring-batch-admin-manager and spring-batch-admin-resources. In addition, override components from Spring Batch Admin

• override-ctx.xml in META-INF/spring/batch/override for setting any additional beans
• web-context.xml in META-INF/spring/batch/servlet/override for the spring beans and Apache FreeMarker templates.
• The jobs are configured in sample-batch-job.xml

The source code is available in GitHub

Implementing the Quartz Job interface

The JobService interface is used for monitoring and management of the batch jobs. For scheduling a job, we will provide an implementation in the application-context with SimpleJobServiceFactoryBean. Refer the override-ctx.xml

Provide an implementation for Quartz Job interface using JobLauncherDetails.java

Scheduling

The Job is scheduled using the Quartz CronScheduleBuilder. The source code is available in GitHub

Persisting scheduled jobs

The Quartz service implementation is storing the information in the domain object BatchJobDataStore only in memory. Hence, the job information is not available once the JVM has been shutdown or restarted. If you have the requirement to persist state to a database, you will need to add custom logic when a job has been scheduled in the QuartzServiceImpl.java. You will also need to load the jobs from the database during the start-up.