spring-batch

This is a spring batch demo application using the org.springframework.boot:spring-boot-starter-batch. The batch is pulling records from a database, process them and insert the processed records back into a new database table.

Description

The project aims to provide a simple example of how to use spring batch to read records from a database table, process them and insert the results into another database table. We will use gradle as our build tool, MySQL as our database.

High-Level Design

1. Input for the batch - the batch will read records from reader table - id, firstName, lastName and random_num.
2. Processing - the batch will process input and join the firstName and lastName to generate the full_name attribute.
3. Output of the batch - the batch will then output and write the data into writer table with attributes - id, full_name and random_num (generated at run time).

Getting Started

You can clone this project git clone https://github.com/geekyjaat/spring-batch.git, run it and expand it further. Or read the tutorial below to start from scratch.

Setting the build tool

we will set up the build.gradle in project directory with plugin java and spring-boot. Also, we will add following dependencies -

1. spring-boot-starter-batch - to pull in spring batch classes.
2. spring-boot-starter-data-jpa - to handle the datasource and pull in spring-jdbc dependencies.
3. mysql-connector-java - java connector for MySQL database.

spring-boot plugin that collects all the jars on the classpath and builds a single, runnable archive jar, which makes it more easier to execute the service.

Here is what our build.gradle looks like -

build.gradle

    group 'com.geeky.batch'
    version '1.0.0'

    apply plugin: 'java'
    apply plugin: 'org.springframework.boot'
    apply plugin: 'io.spring.dependency-management'

    sourceCompatibility = 1.8
    targetCompatibility = 1.8

    jar {
        baseName = 'spring-batch'
    }

    buildscript {
        repositories {
            mavenCentral()
        }
        dependencies {
            classpath('org.springframework.boot:spring-boot-gradle-plugin:2.0.5.RELEASE')
        }
    }

    repositories {
        mavenCentral()
    }

    dependencies {
        compile('org.springframework.boot:spring-boot-starter-batch')
        compile('org.springframework.boot:spring-boot-starter-data-jpa')
        compile('mysql:mysql-connector-java:5.1.38')
        testCompile('junit:junit')
    }

Model

We will add 2 classes to handle our model i.e. input and output data. Here, our input is RecordSO and output is WriterSO. See classes below which we created in package com.barley.batch.model in main/java -

RecordSO

src/main/java/com/barley/batch/model/RecordSO.java

    public class RecordSO {

        private long id;
        private String firstName;
        private String lastName;
        private String randomNum;

        ...getter
        ...setter
    }

src/main/java/com/barley/batch/model/WriterSO.java

WriterSO

    public class WriterSO {

        private long id;
        private String fullName;
        private String randomNum;

        ...getter
        ...setter
    }

Processor

Now we will set up the processor class which will implement spring interface ItemProcessor<I, O>. According to it, you receive an incoming RecordSO object, after which you transform it to a WriterSO.

src/main/java/com/barley/batch/processor/RecordProcessor.java

    public class RecordProcessor implements ItemProcessor<RecordSO, WriterSO> {

        private static final Logger LOGGER = LoggerFactory.getLogger(RecordProcessor.class);

        @Override
        public WriterSO process(RecordSO item) throws Exception {
            LOGGER.info("Processing Record: {}", item);
            WriterSO writerSo = new WriterSO();
            writerSo.setId(item.getId());
            writerSo.setFullName(item.getFirstName() + " " + item.getLastName());
            writerSo.setRandomNum(String.valueOf(Math.random()).substring(3, 8));
            LOGGER.info("Processed Writer: {}", writerSo);
            return writerSo;
        }
    }

Application Properties

Now, we need to set up some spring datasource properties which will help spring configure primary datasource - create application.properties in main/resources -

src/main/resources/application.properties

spring.batch.initialize-schema=ALWAYS
spring.datasource.driverClassName=com.mysql.jdbc.Driver
spring.datasource.url=jdbc:mysql://localhost:3306/batch
spring.datasource.username=root
spring.datasource.password=root
spring.jpa.database-platform=org.hibernate.dialect.MySQLDialect
spring.jpa.show-sql=false
# use below property only if you want spring to use user-configured sql create-drop databases tables. Skip it all tables/data already present.
spring.jpa.hibernate.ddl-auto=create-drop

Primary Database

With spring auto-configuration, the datasource will be configured based application.properties and jars on the classpath. if you are using more than 1, you can set up one of datasources as @Primary @Bean.

Hooking up batch framework

Now, we hook up spring batch components to link up all pieces together - we create the class BatchConfiguration and use this annotations -

1. @Configuration - this tell spring that its a configuration class.
2. @EnableBatchProcessing - this adds a lot of important beans which will support the batch.

    @Configuration
    @EnableBatchProcessing
    public class BatchConfiguration {
    }

ItemReader

Below method creates an ItemReader which will read records from reader table from datasource and turn them into RecordSO.

    @Bean
    public ItemReader<RecordSO> reader() {
        return new JdbcCursorItemReaderBuilder<RecordSO>().name("the-reader")
                .sql("select id, firstName, lastname, random_num from reader").dataSource(dataSource)
                .rowMapper((ResultSet resultSet, int rowNum) -> {
                    if (!(resultSet.isAfterLast()) && !(resultSet.isBeforeFirst())) {
                        RecordSO recordSO = new RecordSO();
                        recordSO.setFirstName(resultSet.getString("firstName"));
                        recordSO.setLastName(resultSet.getString("lastname"));
                        recordSO.setId(resultSet.getInt("Id"));
                        recordSO.setRandomNum(resultSet.getString("random_num"));

                        LOGGER.info("RowMapper record : {}", recordSO);
                        return recordSO;
                    } else {
                        LOGGER.info("Returning null from rowMapper");
                        return null;
                    }
                }).build();
    }

ItemWriter

Below method creates an ItemWriter which will be using the datasouce and will insert the WriterSO into writer table. We are setting up teh sql as well which ItemWriter will use to insert data into table.

    @Bean
    public JdbcBatchItemWriter<WriterSO> writer(DataSource dataSource, ItemPreparedStatementSetter<WriterSO> setter) {
        return new JdbcBatchItemWriterBuilder<WriterSO>()
                .itemSqlParameterSourceProvider(new BeanPropertyItemSqlParameterSourceProvider<>())
                .itemPreparedStatementSetter(setter)
                .sql("insert into writer (id, full_name, random_num) values (?,?,?)").dataSource(dataSource).build();
    }

ItemPreparedStatementSetter

In the writer above, we set a ItemPreparedStatementSetter which will help ItemWriter interpret the sql statement it is using.

    @Bean
    public ItemPreparedStatementSetter<WriterSO> setter() {
        return (item, ps) -> {
            ps.setLong(1, item.getId());
            ps.setString(2, item.getFullName());
            ps.setString(3, item.getRandomNum());
        };
    }

ItemProcessor

Hook up our RecordProcessor in the batch config -

    @Bean
    public ItemProcessor<RecordSO, WriterSO> processor() {
        return new RecordProcessor();
    }

Rest of batch config

Now Spring batch works with concept of jobs, and each job consist of steps where each step can involve a ItemReader, a ItemProcessor and a ItemWriter. So we are going to define a job and set up a step -

    @Bean
    public Job importUserJob(JobCompletionNotificationListener listener, Step step1) {
        return jobBuilderFactory.get("importUserJob").incrementer(new RunIdIncrementer()).listener(listener).flow(step1)
                .end().build();
    }

    @Bean
    public Step step1(JdbcBatchItemWriter<WriterSO> writer, ItemReader<RecordSO> reader) {
        return stepBuilderFactory.get("step1").<RecordSO, WriterSO>chunk(5).reader(reader).processor(processor())
                .writer(writer).build();
    }

Complete code -

src/main/java/com/barley/batch/config/BatchConfiguration.java

Listener

well, we are pretty much ready but we can configure a listener which helps us hook up into spring batch and get to know when the job is finished. This allows us to execute any logic we want to do after completion of batch.

src/main/java/com/barley/batch/listener/JobCompletionNotificationListener.java

    @Component
    public class JobCompletionNotificationListener extends JobExecutionListenerSupport {

        private static final Logger log = LoggerFactory.getLogger(JobCompletionNotificationListener.class);

        @Autowired
        private JdbcTemplate jdbcTemplate;

        @Override
        public void afterJob(JobExecution jobExecution) {
            if (jobExecution.getStatus() == BatchStatus.COMPLETED) {
                log.info("!!! JOB FINISHED! Time to verify the results");

                List<WriterSO> results = jdbcTemplate.query("SELECT id, full_name, random_num FROM writer", (rs, row) -> {
                    WriterSO writerSO = new WriterSO();
                    writerSO.setId(rs.getLong("id"));
                    writerSO.setFullName(rs.getString("full_name"));
                    writerSO.setRandomNum(rs.getString("random_num"));
                    return writerSO;
                });

                for (WriterSO writerSO : results) {
                    log.info("Found <" + writerSO + "> in the database.");
                }
            }
        }
    }

Setting up our database tables -

Now, we will have spring create our database tables each time it runs. You can skip this step if you manually have created the tables or tables with data are already present.

src/main/resources/import.sql

DROP TABLE IF EXISTS reader;
DROP TABLE IF EXISTS writer;

CREATE TABLE `reader` (`id` INT  NOT NULL AUTO_INCREMENT,`firstName`  VARCHAR(20) NULL,`lastName`   VARCHAR(20) NULL,`random_num` VARCHAR(20) NULL, PRIMARY KEY (`id`));

CREATE TABLE `writer` (`id` INT  NOT NULL AUTO_INCREMENT,`full_name`  VARCHAR(40) NULL,`random_num` VARCHAR(20) NULL, PRIMARY KEY (`id`));

INSERT INTO `reader` (`firstName`, `lastName`, `random_num`) VALUES ('abc', 'def', '1');
INSERT INTO `reader` (`firstName`, `lastName`, `random_num`) VALUES ('def', 'zhu', '2');
INSERT INTO `reader` (`firstName`, `lastName`, `random_num`) VALUES ('dummy', 'name', '3');
INSERT INTO `reader` (`firstName`, `lastName`, `random_num`) VALUES ('non', 'pay', '4');
INSERT INTO `reader` (`firstName`, `lastName`, `random_num`) VALUES ('spring', 'batch', '5');

Spring Boot

src/main/java/com/barley/batch/Application.java

    @SpringBootApplication
    public class Application {
        public static void main(String[] args) throws Exception {
            SpringApplication.run(Application.class, args);
        }
    }

Running the app

Done. Let us now run our app -

• Go to terminal and use our build tool to run it - run the cmd -

./gradlew build bootRun.

• ALTERNATIVELY, You can build the jar for the app and then run it - run cmd -

./gradlew build bootRepackage

and then run

java -jar build/libs/spring-batch-1.0.0.jar.

Here is the expected output -

Job: [FlowJob: [name=importUserJob]] launched with the following parameters: [{run.id=1}]
Executing step: [step1]
RowMapper record : RecordSO{id=1, firstName='abc', lastName='def', randomNum='1'}
RowMapper record : RecordSO{id=2, firstName='def', lastName='zhu', randomNum='2'}
RowMapper record : RecordSO{id=3, firstName='dummy', lastName='name', randomNum='3'}
RowMapper record : RecordSO{id=4, firstName='non', lastName='pay', randomNum='4'}
RowMapper record : RecordSO{id=5, firstName='spring', lastName='batch', randomNum='5'}

Processing Record: RecordSO{id=1, firstName='abc', lastName='def', randomNum='1'}
Processed Writer: WriterSO{id=1, fullName='abc def', randomNum='69388'}
Processing Record: RecordSO{id=2, firstName='def', lastName='zhu', randomNum='2'}
Processed Writer: WriterSO{id=2, fullName='def zhu', randomNum='41825'}
Processing Record: RecordSO{id=3, firstName='dummy', lastName='name', randomNum='3'}
Processed Writer: WriterSO{id=3, fullName='dummy name', randomNum='23697'}
Processing Record: RecordSO{id=4, firstName='non', lastName='pay', randomNum='4'}
Processed Writer: WriterSO{id=4, fullName='non pay', randomNum='89450'}
Processing Record: RecordSO{id=5, firstName='spring', lastName='batch', randomNum='5'}
Processed Writer: WriterSO{id=5, fullName='spring batch', randomNum='87917'}

!!! JOB FINISHED! Time to verify the results
Found <WriterSO{id=1, fullName='abc def', randomNum='69388'}> in the database.
Found <WriterSO{id=2, fullName='def zhu', randomNum='41825'}> in the database.
Found <WriterSO{id=3, fullName='dummy name', randomNum='23697'}> in the database.
Found <WriterSO{id=4, fullName='non pay', randomNum='89450'}> in the database.
Found <WriterSO{id=5, fullName='spring batch', randomNum='87917'}> in the database.

Job: [FlowJob: [name=importUserJob]] completed with the following parameters: [{run.id=1}] and the following status: [COMPLETED]

Thank you.