I'm wondering what the best way to load initial database data before the application starts? What I'm looking for is something that will fill my H2 database with data.
For example, I have a domain model "User" I can access users by going to /users but initially there won't be any users in the database so I have to create them. Is there anyway to fill the database with data automatically?
At the moment I have a Bean that gets instantiated by the container and creates users for me.
Example:
@Component
public class DataLoader {
private UserRepository userRepository;
@Autowired
public DataLoader(UserRepository userRepository) {
this.userRepository = userRepository;
LoadUsers();
}
private void LoadUsers() {
userRepository.save(new User("lala", "lala", "lala"));
}
}
But I very much doubt that is the best way of doing it. Or is it?
data.sql
and/or schema.sql
to init data.. All this is documented in the reference guide (which I suggest to read).
You can create a data.sql file in your src/main/resources folder and it will be automatically executed on startup. In this file you can add some insert statements, eg.:
INSERT INTO users (username, firstname, lastname) VALUES
('lala', 'lala', 'lala'),
('lolo', 'lolo', 'lolo');
Similarly, you can create a schema.sql file (or schema-h2.sql) as well to create your schema:
CREATE TABLE task (
id INTEGER PRIMARY KEY,
description VARCHAR(64) NOT NULL,
completed BIT NOT NULL);
Though normally you shouldn't have to do this since Spring boot already configures Hibernate to create your schema based on your entities for an in memory database. If you really want to use schema.sql you'll have to disable this feature by adding this to your application.properties:
spring.jpa.hibernate.ddl-auto=none
More information can be found at the documentation about Database initialization.
If you're using Spring boot 2, database initialization only works for embedded databases (H2, HSQLDB, ...). If you want to use it for other databases as well, you need to change the spring.datasource.initialization-mode
property:
spring.datasource.initialization-mode=always
If you're using multiple database vendors, you can name your file data-h2.sql or data-mysql.sql depending on which database platform you want to use.
To make that work, you'll have to configure the spring.datasource.platform
property though:
spring.datasource.platform=h2
If I just want to insert simple test data I often implement a ApplicationRunner
. Implementations of this interface are run at application startup and can use e.g. a autowired repository to insert some test data.
I think such an implementation would be slightly more explicit than yours because the interface implies that your implementation contains something you would like to do directly after your application is ready.
Your implementation would look sth. like this:
@Component
public class DataLoader implements ApplicationRunner {
private UserRepository userRepository;
@Autowired
public DataLoader(UserRepository userRepository) {
this.userRepository = userRepository;
}
public void run(ApplicationArguments args) {
userRepository.save(new User("lala", "lala", "lala"));
}
}
You can add a spring.datasource.data
property to application.properties
listing the sql files you want to run. Like this:
spring.datasource.data=classpath:accounts.sql, classpath:books.sql, classpath:reviews.sql
//or (depending on SB version)
spring.sql.init.data-locations=classpath:accounts.sql, classpath:books.sql, file:reviews.sql
The sql insert statements in each of these files will then be run, allowing you to keep things tidy.
If you put the files in the classpath, for example in src/main/resources
they will be applied. Or replace classpath:
with file:
and use an absolute path to the file
If you want to run DDL type SQL then use:
spring.datasource.schema=classpath:create_account_table.sql
// depending on spring version
spring.sql.init.schema-locations=classpath:create_account_table.sql
Edit: these solutions are great to get you up and running quickly, however for a more production ready solution it would be worth looking at a framework such as flyway, or liquibase. These frameworks integrate well with spring, and provide a quick, consistent, version-controlled means of initialising schema, and standing-data.
file:
instead of classpath:
.
spring.datasource.schema
for DDL scripts.
spring.sql.init.data-locations
and spring.sql.init.schema-locations
instead.
There are multiple ways how to achieve this. I prefer to use one of following options:
Option 1: Initializing with CommandLineRunner
bean:
@Bean
public CommandLineRunner loadData(CustomerRepository repository) {
return (args) -> {
// save a couple of customers
repository.save(new Customer("Jack", "Bauer"));
repository.save(new Customer("Chloe", "O'Brian"));
repository.save(new Customer("Kim", "Bauer"));
repository.save(new Customer("David", "Palmer"));
repository.save(new Customer("Michelle", "Dessler"));
// fetch all customers
log.info("Customers found with findAll():");
log.info("-------------------------------");
for (Customer customer : repository.findAll()) {
log.info(customer.toString());
}
log.info("");
// fetch an individual customer by ID
Customer customer = repository.findOne(1L);
log.info("Customer found with findOne(1L):");
log.info("--------------------------------");
log.info(customer.toString());
log.info("");
// fetch customers by last name
log.info("Customer found with findByLastNameStartsWithIgnoreCase('Bauer'):");
log.info("--------------------------------------------");
for (Customer bauer : repository
.findByLastNameStartsWithIgnoreCase("Bauer")) {
log.info(bauer.toString());
}
log.info("");
}
}
Option 2: Initializing with schema and data SQL scripts
Prerequisites:
application.properties
spring.jpa.hibernate.ddl-auto=none
Explanation:
Without ddl-auto SQL scripts will be ignored by Hibernate and trigger default behavior - scanning project for @Entity and/or @Table annotated classes.
Then, in your MyApplication
class paste this:
@Bean(name = "dataSource")
public DriverManagerDataSource dataSource() {
DriverManagerDataSource dataSource = new DriverManagerDataSource();
dataSource.setDriverClassName("org.h2.Driver");
dataSource.setUrl("jdbc:h2:~/myDB;MV_STORE=false");
dataSource.setUsername("sa");
dataSource.setPassword("");
// schema init
Resource initSchema = new ClassPathResource("scripts/schema-h2.sql");
Resource initData = new ClassPathResource("scripts/data-h2.sql");
DatabasePopulator databasePopulator = new ResourceDatabasePopulator(initSchema, initData);
DatabasePopulatorUtils.execute(databasePopulator, dataSource);
return dataSource;
}
Where scripts
folder is located under resources
folder (IntelliJ Idea)
Hope it helps someone
Update 04-2021: Both options are great to combine with Spring Profiles as this will help you to avoid creating additional config files making your life as the developer easy.
You can use something like this:
@SpringBootApplication
public class Application {
@Autowired
private UserRepository userRepository;
public static void main(String[] args) {
SpringApplication.run(Application.class, args);
}
@Bean
InitializingBean sendDatabase() {
return () -> {
userRepository.save(new User("John"));
userRepository.save(new User("Rambo"));
};
}
}
In Spring Boot 2 data.sql was not working with me as in spring boot 1.5
import.sql
In addition, a file named import.sql
in the root of the classpath is executed on startup if Hibernate creates the schema from scratch (that is, if the ddl-auto property is set to create or create-drop).
Note very important if you insert Keys cannot be duplicated do not use ddl-auto property is set to update because with each restart will insert same data again
For more information you vist spring websit
https://docs.spring.io/spring-boot/docs/current/reference/html/howto-database-initialization.html
Spring Boot allows you to use a simple script to initialize your database, using Spring Batch.
Still, if you want to use something a bit more elaborated to manage DB versions and so on, Spring Boot integrates well with Flyway.
See also:
Spring Boot Database initialization
You can simply create a import.sql
file in src/main/resources
and Hibernate will execute it when the schema is created.
Here is the way I got that:
@Component
public class ApplicationStartup implements ApplicationListener<ApplicationReadyEvent> {
/**
* This event is executed as late as conceivably possible to indicate that
* the application is ready to service requests.
*/
@Autowired
private MovieRepositoryImpl movieRepository;
@Override
public void onApplicationEvent(final ApplicationReadyEvent event) {
seedData();
}
private void seedData() {
movieRepository.save(new Movie("Example"));
// ... add more code
}
}
Thanks to the author of this article:
http://blog.netgloo.com/2014/11/13/run-code-at-spring-boot-startup/
If you came here and nothing seems to work for you, then it might be the case that you are affected from some changes that were introduced with Spring Boot 2.5
and onwards.
Here is the total set of properties which I use for postgresql.
spring:
sql.init.mode: always <-----------------
datasource:
url: jdbc:postgresql://localhost:5432/products
username:
password:
jpa:
defer-datasource-initialization: true <------------------
hibernate:
ddl-auto: create-drop <----------------
database-platform: org.hibernate.dialect.PostgreSQLDialect
I have also marked with <---
the relevant properties for the current topic in order to achieve the following.
ORM vendor will create database schema for you from Java Entities model.
After database schema is created, initial data will be loaded to database from the file data.sql
Ps: Don't forget to add the file with initial data, data.sql
under src/main/resources
Also as reference: Spring Boot 2.5 release notes
I solved similar problem this way:
@Component
public class DataLoader {
@Autowired
private UserRepository userRepository;
//method invoked during the startup
@PostConstruct
public void loadData() {
userRepository.save(new User("user"));
}
//method invoked during the shutdown
@PreDestroy
public void removeData() {
userRepository.deleteAll();
}
}
You're almost there!
@Component
public class DataLoader implements CommandLineRunner {
private UserRepository userRepository;
public DataLoader(UserRepository userRepository) {
this.userRepository = userRepository;
}
@Override
public void run(String... args) throws Exception {
LoadUsers()
}
private void LoadUsers() {
userRepository.save(new User("lala", "lala", "lala"));
}
}
you can register and event listener to achieve that like below:
@EventListener
public void seed(ContextRefreshedEvent event) {
userRepository.save(new User("lala", "lala", "lala"));
}
When the ContextRefreshEvent is fired, we get access to all autowired beans in the application — including models and repositories.
If someone are struggling in make this to work even following the accepted answer, for me only work adding in my src/test/resources/application.yml
the H2 datasource
details:
spring:
datasource:
platform: h2
url: jdbc:h2:mem:test;DB_CLOSE_DELAY=-1
driver-class-name: org.h2.Driver
username: sa
password:
If you want to insert only few rows and u have JPA Setup. You can use below
@SpringBootApplication
@Slf4j
public class HospitalManagementApplication {
public static void main(String[] args) {
SpringApplication.run(HospitalManagementApplication.class, args);
}
@Bean
ApplicationRunner init(PatientRepository repository) {
return (ApplicationArguments args) -> dataSetup(repository);
}
public void dataSetup(PatientRepository repository){
//inserts
}
You can use the below code. In the following code a database insertion occurs during the startup of the spring boot application.
@SpringBootApplication
public class Application implements CommandLineRunner {
@Autowired
private IService<Car> service;
public static void main(String[] args) {
SpringApplication.run(Application.class, args);
}
@Override
public void run(String... args) throws Exception {
for(int i=1; i<=1000; i++) {
Car car = new Car();
car.setName("Car Name "+i);
book.setPrice(50 + i);
service.saveOrUpdate(car);
}
}
}
This will also work.
@Bean
CommandLineRunner init (StudentRepo studentRepo){
return args -> {
// Adding two students objects
List<String> names = Arrays.asList("udara", "sampath");
names.forEach(name -> studentRepo.save(new Student(name)));
};
}
The most compact (for dynamic data) put @mathias-dpunkt solution into MainApp (with Lombok @AllArgsConstructor
):
@SpringBootApplication
@AllArgsConstructor
public class RestaurantVotingApplication implements ApplicationRunner {
private final VoteRepository voteRepository;
private final UserRepository userRepository;
public static void main(String[] args) {
SpringApplication.run(RestaurantVotingApplication.class, args);
}
@Override
public void run(ApplicationArguments args) {
voteRepository.save(new Vote(userRepository.getOne(1), LocalDate.now(), LocalTime.now()));
}
}
One possibility is using incorrect JDBC URL. make sure it is jdbc:h2:mem:testdb
I created a library that facilitates initial/demo data loading in a Spring Boot application. You can find it at https://github.com/piotrpolak/spring-boot-data-fixtures
Once the data fixtures starter is on the classpath, it will automatically try to load DICTIONARY
data upon application startup (this behavior can be controlled by properties) - all you need to do is to register a bean implementing DataFixture
.
I find loading initial data by code superior to loading it using SQL scripts:
the logic of your fixtures lives close to your application logic/domain model and it is subject to refactoring as your domain evolves
you benefit from incremental demo data updates - imagine a QA environment with some user data (that needs not to be lost after application deploy) but at the same time you want to add data for the new features you developed
Example data fixture:
/**
* You can have as many fixture classes as you want.
* @Order annotation is respected for the fixtures belonging to the same set.
* You can make your demo database to be incrementally updated with fresh data
* each time the application is redeployed - all you need to do is to write
* a good condition in `canBeLoaded()` method.
*/
@Component
public class InitialDataFixture implements DataFixture {
private final LanguageRepository languageRepository;
// ...
@Override
public DataFixtureSet getSet() {
return DataFixtureSet.DICTIONARY;
}
/**
* We want to make sure the fixture is applied once and once only.
* A more sophisticated condition can be used to create incremental demo data
* over time without the need to reset the QA database (for example).
*/
@Override
public boolean canBeLoaded() {
return languageRepository.size() == 0;
}
/**
* The actual application of the fixture.
* Assuming that data fixtures are registered as beans, this method can call
* other services and/or repositories.
*/
@Override
public void load() {
languageRepository.saveAll(Arrays.asList(
new Language("en-US"), new Language("pl-PL")));
}
}
The concept is inspired by the Symfony Doctrine Data Fixtures bundle.
canBeLoaded()
method and then you might as well have gone with Liquibase or Flyway. One of the troubles that Liquibase solves for you is concurrency: What happens if your application is multi-node and several instances of your application start at the same time? And what about rollback if something unforseen happens? In what state is you db left then? These are hard problems to tackle and both Liquibase and Flyway have spent years ironing out the edges.
For those using MysqlDriver, I tried using Init attribute of @bean annotation and it works.
After created the Schema and Data sql file in the path of resources\Scripts
Add the line in application.properties
spring.jpa.hibernate.ddl-auto=none
Edit the Application content:
package com.spring_mvaen.demo;
import org.springframework.boot.CommandLineRunner;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.context.annotation.Bean;
import org.springframework.core.io.ClassPathResource;
import org.springframework.core.io.Resource;
import org.springframework.jdbc.datasource.DriverManagerDataSource;
import org.springframework.jdbc.datasource.init.DatabasePopulator;
import org.springframework.jdbc.datasource.init.DatabasePopulatorUtils;
import org.springframework.jdbc.datasource.init.ResourceDatabasePopulator;
@SpringBootApplication
public class DemoApplication implements CommandLineRunner {
public static void main(String[] args) {
SpringApplication.run(DemoApplication.class, args);
}
@Override
public void run(String... arg0) throws Exception {
System.out.println("Hello world from Command Line Runner");
}
@Bean(name = "dataSource")
public DriverManagerDataSource dataSource() {
DriverManagerDataSource dataSource = new DriverManagerDataSource();
dataSource.setDriverClassName("com.mysql.cj.jdbc.Driver");
dataSource.setUrl("jdbc:mysql://localhost:3306/db_spring_rest?useUnicode=true&useLegacyDatetimeCode=fa lse&serverTimezone=UTC&createDatabaseIfNotExist=true&allowPublicKeyRetrieval=true&useSSL=false");
dataSource.setUsername("root");
dataSource.setPassword("root");
// schema init
Resource initSchema = new ClassPathResource("scripts/schema.sql");
Resource initData = new ClassPathResource("scripts/data.sql");
DatabasePopulator databasePopulator = new ResourceDatabasePopulator(initSchema, initData);
DatabasePopulatorUtils.execute(databasePopulator, dataSource);
return dataSource;
}
}
Success story sharing
schema.sql
/data.sql
files will be executed whenspring.datasource.initialize
istrue
(which is the default).spring.jpa.hibernate.ddl-auto
can be used to generate your tables based on your entity configuration rather than using a SQL file. This is by default enabled on in-memory databases. That's why I added the note in my answer, explaining that if you use an in-memory database and you want to use theschema.sql
, you need to disablespring.jpa.hibernate.ddl-auto
otherwise both will try to create your table.data-h2.sql
filename for your initial data, you should setspring.datasource.platform=h2
in your application properties as well.org.h2.jdbc.JdbcSQLException
-exception, because the data is already present in the database. I am using an embedded H2 database, but the problem stays the same.MERGE INTO
. I figured out, that there is a way to circumvent this using an import.sql file instead of a data.sql. It requiresspring.jpa.hibernate.ddl-auto
to create or create-drop. Then whenever the schema file is created (and/or a schema.sql is executed), the import.sql is executed as well. Still: it feels like a workaround and not a clean implementation of creating init data.