This section goes into more detail about how you should use Spring Boot. It covers topics such as build systems, auto-configuration, and how to run your applications. We also cover some Spring Boot best practices. Although there is nothing particularly special about Spring Boot (it is just another library that you can consume), there are a few recommendations that, when followed, make your development process a little easier.

If you are starting out with Spring Boot, you should probably read the Getting Started guide before diving into this section.

1. Build Systems

It is strongly recommended that you choose a build system that supports dependency management and that can consume artifacts published to the “Maven Central” repository. We would recommend that you choose Maven or Gradle. It is possible to get Spring Boot to work with other build systems (Ant, for example), but they are not particularly well supported.

1.1. Dependency Management

Each release of Spring Boot provides a curated list of dependencies that it supports. In practice, you do not need to provide a version for any of these dependencies in your build configuration, as Spring Boot manages that for you. When you upgrade Spring Boot itself, these dependencies are upgraded as well in a consistent way.

You can still specify a version and override Spring Boot’s recommendations if you need to do so.

The curated list contains all the Spring modules that you can use with Spring Boot as well as a refined list of third party libraries. The list is available as a standard Bills of Materials (spring-boot-dependencies) that can be used with both Maven and Gradle.

Each release of Spring Boot is associated with a base version of the Spring Framework. We highly recommend that you do not specify its version.

1.2. Maven

To learn about using Spring Boot with Maven, see the documentation for Spring Boot’s Maven plugin:

1.3. Gradle

To learn about using Spring Boot with Gradle, see the documentation for Spring Boot’s Gradle plugin:

1.4. Ant

It is possible to build a Spring Boot project using Apache Ant+Ivy. The spring-boot-antlib “AntLib” module is also available to help Ant create executable jars.

To declare dependencies, a typical ivy.xml file looks something like the following example:

<ivy-module version="2.0">
    <info organisation="org.springframework.boot" module="spring-boot-sample-ant" />
    <configurations>
        <conf name="compile" description="everything needed to compile this module" />
        <conf name="runtime" extends="compile" description="everything needed to run this module" />
    </configurations>
    <dependencies>
        <dependency org="org.springframework.boot" name="spring-boot-starter"
            rev="${spring-boot.version}" conf="compile" />
    </dependencies>
</ivy-module>

A typical build.xml looks like the following example:

<project
    xmlns:ivy="antlib:org.apache.ivy.ant"
    xmlns:spring-boot="antlib:org.springframework.boot.ant"
    name="myapp" default="build">

    <property name="spring-boot.version" value="3.2.10" />

    <target name="resolve" description="--> retrieve dependencies with ivy">
        <ivy:retrieve pattern="lib/[conf]/[artifact]-[type]-[revision].[ext]" />
    </target>

    <target name="classpaths" depends="resolve">
        <path id="compile.classpath">
            <fileset dir="lib/compile" includes="*.jar" />
        </path>
    </target>

    <target name="init" depends="classpaths">
        <mkdir dir="build/classes" />
    </target>

    <target name="compile" depends="init" description="compile">
        <javac srcdir="src/main/java" destdir="build/classes" classpathref="compile.classpath" />
    </target>

    <target name="build" depends="compile">
        <spring-boot:exejar destfile="build/myapp.jar" classes="build/classes">
            <spring-boot:lib>
                <fileset dir="lib/runtime" />
            </spring-boot:lib>
        </spring-boot:exejar>
    </target>
</project>
If you do not want to use the spring-boot-antlib module, see the howto.html “How-to” .

1.5. Starters

Starters are a set of convenient dependency descriptors that you can include in your application. You get a one-stop shop for all the Spring and related technologies that you need without having to hunt through sample code and copy-paste loads of dependency descriptors. For example, if you want to get started using Spring and JPA for database access, include the spring-boot-starter-data-jpa dependency in your project.

The starters contain a lot of the dependencies that you need to get a project up and running quickly and with a consistent, supported set of managed transitive dependencies.

What is in a name

All official starters follow a similar naming pattern; spring-boot-starter-*, where * is a particular type of application. This naming structure is intended to help when you need to find a starter. The Maven integration in many IDEs lets you search dependencies by name. For example, with the appropriate Eclipse or Spring Tools plugin installed, you can press ctrl-space in the POM editor and type “spring-boot-starter” for a complete list.

As explained in the “Creating Your Own Starter” section, third party starters should not start with spring-boot, as it is reserved for official Spring Boot artifacts. Rather, a third-party starter typically starts with the name of the project. For example, a third-party starter project called thirdpartyproject would typically be named thirdpartyproject-spring-boot-starter.

The following application starters are provided by Spring Boot under the org.springframework.boot group:

Table 1. Spring Boot application starters
Name Description

spring-boot-starter

Core starter, including auto-configuration support, logging and YAML

spring-boot-starter-activemq

Starter for JMS messaging using Apache ActiveMQ

spring-boot-starter-amqp

Starter for using Spring AMQP and Rabbit MQ

spring-boot-starter-aop

Starter for aspect-oriented programming with Spring AOP and AspectJ

spring-boot-starter-artemis

Starter for JMS messaging using Apache Artemis

spring-boot-starter-batch

Starter for using Spring Batch

spring-boot-starter-cache

Starter for using Spring Framework’s caching support

spring-boot-starter-data-cassandra

Starter for using Cassandra distributed database and Spring Data Cassandra

spring-boot-starter-data-cassandra-reactive

Starter for using Cassandra distributed database and Spring Data Cassandra Reactive

spring-boot-starter-data-couchbase

Starter for using Couchbase document-oriented database and Spring Data Couchbase

spring-boot-starter-data-couchbase-reactive

Starter for using Couchbase document-oriented database and Spring Data Couchbase Reactive

spring-boot-starter-data-elasticsearch

Starter for using Elasticsearch search and analytics engine and Spring Data Elasticsearch

spring-boot-starter-data-jdbc

Starter for using Spring Data JDBC

spring-boot-starter-data-jpa

Starter for using Spring Data JPA with Hibernate

spring-boot-starter-data-ldap

Starter for using Spring Data LDAP

spring-boot-starter-data-mongodb

Starter for using MongoDB document-oriented database and Spring Data MongoDB

spring-boot-starter-data-mongodb-reactive

Starter for using MongoDB document-oriented database and Spring Data MongoDB Reactive

spring-boot-starter-data-neo4j

Starter for using Neo4j graph database and Spring Data Neo4j

spring-boot-starter-data-r2dbc

Starter for using Spring Data R2DBC

spring-boot-starter-data-redis

Starter for using Redis key-value data store with Spring Data Redis and the Lettuce client

spring-boot-starter-data-redis-reactive

Starter for using Redis key-value data store with Spring Data Redis reactive and the Lettuce client

spring-boot-starter-data-rest

Starter for exposing Spring Data repositories over REST using Spring Data REST and Spring MVC

spring-boot-starter-freemarker

Starter for building MVC web applications using FreeMarker views

spring-boot-starter-graphql

Starter for building GraphQL applications with Spring GraphQL

spring-boot-starter-groovy-templates

Starter for building MVC web applications using Groovy Templates views

spring-boot-starter-hateoas

Starter for building hypermedia-based RESTful web application with Spring MVC and Spring HATEOAS

spring-boot-starter-integration

Starter for using Spring Integration

spring-boot-starter-jdbc

Starter for using JDBC with the HikariCP connection pool

spring-boot-starter-jersey

Starter for building RESTful web applications using JAX-RS and Jersey. An alternative to spring-boot-starter-web

spring-boot-starter-jooq

Starter for using jOOQ to access SQL databases with JDBC. An alternative to spring-boot-starter-data-jpa or spring-boot-starter-jdbc

spring-boot-starter-json

Starter for reading and writing json

spring-boot-starter-mail

Starter for using Java Mail and Spring Framework’s email sending support

spring-boot-starter-mustache

Starter for building web applications using Mustache views

spring-boot-starter-oauth2-authorization-server

Starter for using Spring Authorization Server features

spring-boot-starter-oauth2-client

Starter for using Spring Security’s OAuth2/OpenID Connect client features

spring-boot-starter-oauth2-resource-server

Starter for using Spring Security’s OAuth2 resource server features

spring-boot-starter-pulsar

Starter for using Spring for Apache Pulsar

spring-boot-starter-pulsar-reactive

Starter for using Spring for Apache Pulsar Reactive

spring-boot-starter-quartz

Starter for using the Quartz scheduler

spring-boot-starter-rsocket

Starter for building RSocket clients and servers

spring-boot-starter-security

Starter for using Spring Security

spring-boot-starter-test

Starter for testing Spring Boot applications with libraries including JUnit Jupiter, Hamcrest and Mockito

spring-boot-starter-thymeleaf

Starter for building MVC web applications using Thymeleaf views

spring-boot-starter-validation

Starter for using Java Bean Validation with Hibernate Validator

spring-boot-starter-web

Starter for building web, including RESTful, applications using Spring MVC. Uses Tomcat as the default embedded container

spring-boot-starter-web-services

Starter for using Spring Web Services

spring-boot-starter-webflux

Starter for building WebFlux applications using Spring Framework’s Reactive Web support

spring-boot-starter-websocket

Starter for building WebSocket applications using Spring Framework’s MVC WebSocket support

In addition to the application starters, the following starters can be used to add production ready features:

Table 2. Spring Boot production starters
Name Description

spring-boot-starter-actuator

Starter for using Spring Boot’s Actuator which provides production ready features to help you monitor and manage your application

Finally, Spring Boot also includes the following starters that can be used if you want to exclude or swap specific technical facets:

Table 3. Spring Boot technical starters
Name Description

spring-boot-starter-jetty

Starter for using Jetty as the embedded servlet container. An alternative to spring-boot-starter-tomcat

spring-boot-starter-log4j2

Starter for using Log4j2 for logging. An alternative to spring-boot-starter-logging

spring-boot-starter-logging

Starter for logging using Logback. Default logging starter

spring-boot-starter-reactor-netty

Starter for using Reactor Netty as the embedded reactive HTTP server.

spring-boot-starter-tomcat

Starter for using Tomcat as the embedded servlet container. Default servlet container starter used by spring-boot-starter-web

spring-boot-starter-undertow

Starter for using Undertow as the embedded servlet container. An alternative to spring-boot-starter-tomcat

To learn how to swap technical facets, please see the how-to documentation for swapping web server and logging system.

For a list of additional community contributed starters, see the README file in the spring-boot-starters module on GitHub.

2. Structuring Your Code

Spring Boot does not require any specific code layout to work. However, there are some best practices that help.

If you wish to enforce a structure based on domains, take a look at Spring Modulith.

2.1. Using the “default” Package

When a class does not include a package declaration, it is considered to be in the “default package”. The use of the “default package” is generally discouraged and should be avoided. It can cause particular problems for Spring Boot applications that use the @ComponentScan, @ConfigurationPropertiesScan, @EntityScan, or @SpringBootApplication annotations, since every class from every jar is read.

We recommend that you follow Java’s recommended package naming conventions and use a reversed domain name (for example, com.example.project).

2.2. Locating the Main Application Class

We generally recommend that you locate your main application class in a root package above other classes. The @SpringBootApplication annotation is often placed on your main class, and it implicitly defines a base “search package” for certain items. For example, if you are writing a JPA application, the package of the @SpringBootApplication annotated class is used to search for @Entity items. Using a root package also allows component scan to apply only on your project.

If you do not want to use @SpringBootApplication, the @EnableAutoConfiguration and @ComponentScan annotations that it imports defines that behavior so you can also use those instead.

The following listing shows a typical layout:

com
 +- example
     +- myapplication
         +- MyApplication.java
         |
         +- customer
         |   +- Customer.java
         |   +- CustomerController.java
         |   +- CustomerService.java
         |   +- CustomerRepository.java
         |
         +- order
             +- Order.java
             +- OrderController.java
             +- OrderService.java
             +- OrderRepository.java

The MyApplication.java file would declare the main method, along with the basic @SpringBootApplication, as follows:

Java
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;

@SpringBootApplication
public class MyApplication {

    public static void main(String[] args) {
        SpringApplication.run(MyApplication.class, args);
    }

}
Kotlin
import org.springframework.boot.autoconfigure.SpringBootApplication
import org.springframework.boot.runApplication

@SpringBootApplication
class MyApplication

fun main(args: Array<String>) {
    runApplication<MyApplication>(*args)
}

3. Configuration Classes

Spring Boot favors Java-based configuration. Although it is possible to use SpringApplication with XML sources, we generally recommend that your primary source be a single @Configuration class. Usually the class that defines the main method is a good candidate as the primary @Configuration.

Many Spring configuration examples have been published on the Internet that use XML configuration. If possible, always try to use the equivalent Java-based configuration. Searching for Enable* annotations can be a good starting point.

3.1. Importing Additional Configuration Classes

You need not put all your @Configuration into a single class. The @Import annotation can be used to import additional configuration classes. Alternatively, you can use @ComponentScan to automatically pick up all Spring components, including @Configuration classes.

3.2. Importing XML Configuration

If you absolutely must use XML based configuration, we recommend that you still start with a @Configuration class. You can then use an @ImportResource annotation to load XML configuration files.

4. Auto-configuration

Spring Boot auto-configuration attempts to automatically configure your Spring application based on the jar dependencies that you have added. For example, if HSQLDB is on your classpath, and you have not manually configured any database connection beans, then Spring Boot auto-configures an in-memory database.

You need to opt-in to auto-configuration by adding the @EnableAutoConfiguration or @SpringBootApplication annotations to one of your @Configuration classes.

You should only ever add one @SpringBootApplication or @EnableAutoConfiguration annotation. We generally recommend that you add one or the other to your primary @Configuration class only.

4.1. Gradually Replacing Auto-configuration

Auto-configuration is non-invasive. At any point, you can start to define your own configuration to replace specific parts of the auto-configuration. For example, if you add your own DataSource bean, the default embedded database support backs away.

If you need to find out what auto-configuration is currently being applied, and why, start your application with the --debug switch. Doing so enables debug logs for a selection of core loggers and logs a conditions report to the console.

4.2. Disabling Specific Auto-configuration Classes

If you find that specific auto-configuration classes that you do not want are being applied, you can use the exclude attribute of @SpringBootApplication to disable them, as shown in the following example:

Java
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.boot.autoconfigure.jdbc.DataSourceAutoConfiguration;

@SpringBootApplication(exclude = { DataSourceAutoConfiguration.class })
public class MyApplication {

}
Kotlin
import org.springframework.boot.autoconfigure.SpringBootApplication
import org.springframework.boot.autoconfigure.jdbc.DataSourceAutoConfiguration

@SpringBootApplication(exclude = [DataSourceAutoConfiguration::class])
class MyApplication

If the class is not on the classpath, you can use the excludeName attribute of the annotation and specify the fully qualified name instead. If you prefer to use @EnableAutoConfiguration rather than @SpringBootApplication, exclude and excludeName are also available. Finally, you can also control the list of auto-configuration classes to exclude by using the spring.autoconfigure.exclude property.

You can define exclusions both at the annotation level and by using the property.
Even though auto-configuration classes are public, the only aspect of the class that is considered public API is the name of the class which can be used for disabling the auto-configuration. The actual contents of those classes, such as nested configuration classes or bean methods are for internal use only and we do not recommend using those directly.

4.3. Auto-configuration Packages

Auto-configuration packages are the packages that various auto-configured features look in by default when scanning for things such as entities and Spring Data repositories. The @EnableAutoConfiguration annotation (either directly or through its presence on @SpringBootApplication) determines the default auto-configuration package. Additional packages can be configured using the @AutoConfigurationPackage annotation.

5. Spring Beans and Dependency Injection

You are free to use any of the standard Spring Framework techniques to define your beans and their injected dependencies. We generally recommend using constructor injection to wire up dependencies and @ComponentScan to find beans.

If you structure your code as suggested above (locating your application class in a top package), you can add @ComponentScan without any arguments or use the @SpringBootApplication annotation which implicitly includes it. All of your application components (@Component, @Service, @Repository, @Controller, and others) are automatically registered as Spring Beans.

The following example shows a @Service Bean that uses constructor injection to obtain a required RiskAssessor bean:

Java
import org.springframework.stereotype.Service;

@Service
public class MyAccountService implements AccountService {

    private final RiskAssessor riskAssessor;

    public MyAccountService(RiskAssessor riskAssessor) {
        this.riskAssessor = riskAssessor;
    }

    // ...

}
Kotlin
import org.springframework.stereotype.Service

@Service
class MyAccountService(private val riskAssessor: RiskAssessor) : AccountService

If a bean has more than one constructor, you will need to mark the one you want Spring to use with @Autowired:

Java
import java.io.PrintStream;

import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Service;

@Service
public class MyAccountService implements AccountService {

    private final RiskAssessor riskAssessor;

    private final PrintStream out;

    @Autowired
    public MyAccountService(RiskAssessor riskAssessor) {
        this.riskAssessor = riskAssessor;
        this.out = System.out;
    }

    public MyAccountService(RiskAssessor riskAssessor, PrintStream out) {
        this.riskAssessor = riskAssessor;
        this.out = out;
    }

    // ...

}
Kotlin
import org.springframework.beans.factory.annotation.Autowired
import org.springframework.stereotype.Service
import java.io.PrintStream

@Service
class MyAccountService : AccountService {

    private val riskAssessor: RiskAssessor

    private val out: PrintStream

    @Autowired
    constructor(riskAssessor: RiskAssessor) {
        this.riskAssessor = riskAssessor
        out = System.out
    }

    constructor(riskAssessor: RiskAssessor, out: PrintStream) {
        this.riskAssessor = riskAssessor
        this.out = out
    }

    // ...

}
Notice how using constructor injection lets the riskAssessor field be marked as final, indicating that it cannot be subsequently changed.

6. Using the @SpringBootApplication Annotation

Many Spring Boot developers like their apps to use auto-configuration, component scan and be able to define extra configuration on their "application class". A single @SpringBootApplication annotation can be used to enable those three features, that is:

  • @EnableAutoConfiguration: enable Spring Boot’s auto-configuration mechanism

  • @ComponentScan: enable @Component scan on the package where the application is located (see the best practices)

  • @SpringBootConfiguration: enable registration of extra beans in the context or the import of additional configuration classes. An alternative to Spring’s standard @Configuration that aids configuration detection in your integration tests.

Java
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;

// Same as @SpringBootConfiguration @EnableAutoConfiguration @ComponentScan
@SpringBootApplication
public class MyApplication {

    public static void main(String[] args) {
        SpringApplication.run(MyApplication.class, args);
    }

}
Kotlin
import org.springframework.boot.autoconfigure.SpringBootApplication
import org.springframework.boot.runApplication

// same as @SpringBootConfiguration @EnableAutoConfiguration @ComponentScan
@SpringBootApplication
class MyApplication

fun main(args: Array<String>) {
    runApplication<MyApplication>(*args)
}
@SpringBootApplication also provides aliases to customize the attributes of @EnableAutoConfiguration and @ComponentScan.

None of these features are mandatory and you may choose to replace this single annotation by any of the features that it enables. For instance, you may not want to use component scan or configuration properties scan in your application:

Java
import org.springframework.boot.SpringApplication;
import org.springframework.boot.SpringBootConfiguration;
import org.springframework.boot.autoconfigure.EnableAutoConfiguration;
import org.springframework.context.annotation.Import;

@SpringBootConfiguration(proxyBeanMethods = false)
@EnableAutoConfiguration
@Import({ SomeConfiguration.class, AnotherConfiguration.class })
public class MyApplication {

    public static void main(String[] args) {
        SpringApplication.run(MyApplication.class, args);
    }

}
Kotlin
import org.springframework.boot.SpringBootConfiguration
import org.springframework.boot.autoconfigure.EnableAutoConfiguration
import org.springframework.boot.docs.using.structuringyourcode.locatingthemainclass.MyApplication
import org.springframework.boot.runApplication
import org.springframework.context.annotation.Import

@SpringBootConfiguration(proxyBeanMethods = false)
@EnableAutoConfiguration
@Import(SomeConfiguration::class, AnotherConfiguration::class)
class MyApplication

fun main(args: Array<String>) {
    runApplication<MyApplication>(*args)
}

In this example, MyApplication is just like any other Spring Boot application except that @Component-annotated classes and @ConfigurationProperties-annotated classes are not detected automatically and the user-defined beans are imported explicitly (see @Import).

7. Running Your Application

One of the biggest advantages of packaging your application as a jar and using an embedded HTTP server is that you can run your application as you would any other. The sample applies to debugging Spring Boot applications. You do not need any special IDE plugins or extensions.

This section only covers jar-based packaging. If you choose to package your application as a war file, see your server and IDE documentation.

7.1. Running From an IDE

You can run a Spring Boot application from your IDE as a Java application. However, you first need to import your project. Import steps vary depending on your IDE and build system. Most IDEs can import Maven projects directly. For example, Eclipse users can select Import…​Existing Maven Projects from the File menu.

If you cannot directly import your project into your IDE, you may be able to generate IDE metadata by using a build plugin. Maven includes plugins for Eclipse and IDEA. Gradle offers plugins for various IDEs.

If you accidentally run a web application twice, you see a “Port already in use” error. Spring Tools users can use the Relaunch button rather than the Run button to ensure that any existing instance is closed.

7.2. Running as a Packaged Application

If you use the Spring Boot Maven or Gradle plugins to create an executable jar, you can run your application using java -jar, as shown in the following example:

$ java -jar target/myapplication-0.0.1-SNAPSHOT.jar

It is also possible to run a packaged application with remote debugging support enabled. Doing so lets you attach a debugger to your packaged application, as shown in the following example:

$ java -agentlib:jdwp=server=y,transport=dt_socket,address=8000,suspend=n \
       -jar target/myapplication-0.0.1-SNAPSHOT.jar

7.3. Using the Maven Plugin

The Spring Boot Maven plugin includes a run goal that can be used to quickly compile and run your application. Applications run in an exploded form, as they do in your IDE. The following example shows a typical Maven command to run a Spring Boot application:

$ mvn spring-boot:run

You might also want to use the MAVEN_OPTS operating system environment variable, as shown in the following example:

$ export MAVEN_OPTS=-Xmx1024m

7.4. Using the Gradle Plugin

The Spring Boot Gradle plugin also includes a bootRun task that can be used to run your application in an exploded form. The bootRun task is added whenever you apply the org.springframework.boot and java plugins and is shown in the following example:

$ gradle bootRun

You might also want to use the JAVA_OPTS operating system environment variable, as shown in the following example:

$ export JAVA_OPTS=-Xmx1024m

7.5. Hot Swapping

Since Spring Boot applications are plain Java applications, JVM hot-swapping should work out of the box. JVM hot swapping is somewhat limited with the bytecode that it can replace. For a more complete solution, JRebel can be used.

The spring-boot-devtools module also includes support for quick application restarts. See the Hot swapping “How-to” for details.

8. Developer Tools

Spring Boot includes an additional set of tools that can make the application development experience a little more pleasant. The spring-boot-devtools module can be included in any project to provide additional development-time features. To include devtools support, add the module dependency to your build, as shown in the following listings for Maven and Gradle:

Maven
<dependencies>
    <dependency>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-devtools</artifactId>
        <optional>true</optional>
    </dependency>
</dependencies>
Gradle
dependencies {
    developmentOnly("org.springframework.boot:spring-boot-devtools")
}
Devtools might cause classloading issues, in particular in multi-module projects. Diagnosing Classloading Issues explains how to diagnose and solve them.
Developer tools are automatically disabled when running a fully packaged application. If your application is launched from java -jar or if it is started from a special classloader, then it is considered a “production application”. You can control this behavior by using the spring.devtools.restart.enabled system property. To enable devtools, irrespective of the classloader used to launch your application, set the -Dspring.devtools.restart.enabled=true system property. This must not be done in a production environment where running devtools is a security risk. To disable devtools, exclude the dependency or set the -Dspring.devtools.restart.enabled=false system property.
Flagging the dependency as optional in Maven or using the developmentOnly configuration in Gradle (as shown above) prevents devtools from being transitively applied to other modules that use your project.
Repackaged archives do not contain devtools by default. If you want to use a certain remote devtools feature, you need to include it. When using the Maven plugin, set the excludeDevtools property to false. When using the Gradle plugin, configure the task’s classpath to include the developmentOnly configuration.

8.1. Diagnosing Classloading Issues

As described in the Restart vs Reload section, restart functionality is implemented by using two classloaders. For most applications, this approach works well. However, it can sometimes cause classloading issues, in particular in multi-module projects.

To diagnose whether the classloading issues are indeed caused by devtools and its two classloaders, try disabling restart. If this solves your problems, customize the restart classloader to include your entire project.

8.2. Property Defaults

Several of the libraries supported by Spring Boot use caches to improve performance. For example, template engines cache compiled templates to avoid repeatedly parsing template files. Also, Spring MVC can add HTTP caching headers to responses when serving static resources.

While caching is very beneficial in production, it can be counter-productive during development, preventing you from seeing the changes you just made in your application. For this reason, spring-boot-devtools disables the caching options by default.

Cache options are usually configured by settings in your application.properties file. For example, Thymeleaf offers the spring.thymeleaf.cache property. Rather than needing to set these properties manually, the spring-boot-devtools module automatically applies sensible development-time configuration.

The following table lists all the properties that are applied:

Name Default Value

server.error.include-binding-errors

always

server.error.include-message

always

server.error.include-stacktrace

always

server.servlet.jsp.init-parameters.development

true

server.servlet.session.persistent

true

spring.docker.compose.readiness.wait

only-if-started

spring.freemarker.cache

false

spring.graphql.graphiql.enabled

true

spring.groovy.template.cache

false

spring.h2.console.enabled

true

spring.mustache.servlet.cache

false

spring.mvc.log-resolved-exception

true

spring.reactor.netty.shutdown-quiet-period

0s

spring.template.provider.cache

false

spring.thymeleaf.cache

false

spring.web.resources.cache.period

0

spring.web.resources.chain.cache

false

If you do not want property defaults to be applied you can set spring.devtools.add-properties to false in your application.properties.

Because you need more information about web requests while developing Spring MVC and Spring WebFlux applications, developer tools suggests you to enable DEBUG logging for the web logging group. This will give you information about the incoming request, which handler is processing it, the response outcome, and other details. If you wish to log all request details (including potentially sensitive information), you can turn on the spring.mvc.log-request-details or spring.codec.log-request-details configuration properties.

8.3. Automatic Restart

Applications that use spring-boot-devtools automatically restart whenever files on the classpath change. This can be a useful feature when working in an IDE, as it gives a very fast feedback loop for code changes. By default, any entry on the classpath that points to a directory is monitored for changes. Note that certain resources, such as static assets and view templates, do not need to restart the application.

Triggering a restart

As DevTools monitors classpath resources, the only way to trigger a restart is to update the classpath. Whether you’re using an IDE or one of the build plugins, the modified files have to be recompiled to trigger a restart. The way in which you cause the classpath to be updated depends on the tool that you are using:

  • In Eclipse, saving a modified file causes the classpath to be updated and triggers a restart.

  • In IntelliJ IDEA, building the project (Build +→+ Build Project) has the same effect.

  • If using a build plugin, running mvn compile for Maven or gradle build for Gradle will trigger a restart.

If you are restarting with Maven or Gradle using the build plugin you must leave the forking set to enabled. If you disable forking, the isolated application classloader used by devtools will not be created and restarts will not operate properly.
Automatic restart works very well when used with LiveReload. See the LiveReload section for details. If you use JRebel, automatic restarts are disabled in favor of dynamic class reloading. Other devtools features (such as LiveReload and property overrides) can still be used.
DevTools relies on the application context’s shutdown hook to close it during a restart. It does not work correctly if you have disabled the shutdown hook (SpringApplication.setRegisterShutdownHook(false)).
DevTools needs to customize the ResourceLoader used by the ApplicationContext. If your application provides one already, it is going to be wrapped. Direct override of the getResource method on the ApplicationContext is not supported.
Automatic restart is not supported when using AspectJ weaving.
Restart vs Reload

The restart technology provided by Spring Boot works by using two classloaders. Classes that do not change (for example, those from third-party jars) are loaded into a base classloader. Classes that you are actively developing are loaded into a restart classloader. When the application is restarted, the restart classloader is thrown away and a new one is created. This approach means that application restarts are typically much faster than “cold starts”, since the base classloader is already available and populated.

If you find that restarts are not quick enough for your applications or you encounter classloading issues, you could consider reloading technologies such as JRebel from ZeroTurnaround. These work by rewriting classes as they are loaded to make them more amenable to reloading.

8.3.1. Logging Changes in Condition Evaluation

By default, each time your application restarts, a report showing the condition evaluation delta is logged. The report shows the changes to your application’s auto-configuration as you make changes such as adding or removing beans and setting configuration properties.

To disable the logging of the report, set the following property:

Properties
spring.devtools.restart.log-condition-evaluation-delta=false
Yaml
spring:
  devtools:
    restart:
      log-condition-evaluation-delta: false

8.3.2. Excluding Resources

Certain resources do not necessarily need to trigger a restart when they are changed. For example, Thymeleaf templates can be edited in-place. By default, changing resources in /META-INF/maven, /META-INF/resources, /resources, /static, /public, or /templates does not trigger a restart but does trigger a live reload. If you want to customize these exclusions, you can use the spring.devtools.restart.exclude property. For example, to exclude only /static and /public you would set the following property:

Properties
spring.devtools.restart.exclude=static/**,public/**
Yaml
spring:
  devtools:
    restart:
      exclude: "static/**,public/**"
If you want to keep those defaults and add additional exclusions, use the spring.devtools.restart.additional-exclude property instead.

8.3.3. Watching Additional Paths

You may want your application to be restarted or reloaded when you make changes to files that are not on the classpath. To do so, use the spring.devtools.restart.additional-paths property to configure additional paths to watch for changes. You can use the spring.devtools.restart.exclude property described earlier to control whether changes beneath the additional paths trigger a full restart or a live reload.

8.3.4. Disabling Restart

If you do not want to use the restart feature, you can disable it by using the spring.devtools.restart.enabled property. In most cases, you can set this property in your application.properties (doing so still initializes the restart classloader, but it does not watch for file changes).

If you need to completely disable restart support (for example, because it does not work with a specific library), you need to set the spring.devtools.restart.enabled System property to false before calling SpringApplication.run(…​), as shown in the following example:

Java
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;

@SpringBootApplication
public class MyApplication {

    public static void main(String[] args) {
        System.setProperty("spring.devtools.restart.enabled", "false");
        SpringApplication.run(MyApplication.class, args);
    }

}
Kotlin
import org.springframework.boot.SpringApplication
import org.springframework.boot.autoconfigure.SpringBootApplication

@SpringBootApplication
object MyApplication {

    @JvmStatic
    fun main(args: Array<String>) {
        System.setProperty("spring.devtools.restart.enabled", "false")
        SpringApplication.run(MyApplication::class.java, *args)
    }

}

8.3.5. Using a Trigger File

If you work with an IDE that continuously compiles changed files, you might prefer to trigger restarts only at specific times. To do so, you can use a “trigger file”, which is a special file that must be modified when you want to actually trigger a restart check.

Any update to the file will trigger a check, but restart only actually occurs if Devtools has detected it has something to do.

To use a trigger file, set the spring.devtools.restart.trigger-file property to the name (excluding any path) of your trigger file. The trigger file must appear somewhere on your classpath.

For example, if you have a project with the following structure:

src
+- main
   +- resources
      +- .reloadtrigger

Then your trigger-file property would be:

Properties
spring.devtools.restart.trigger-file=.reloadtrigger
Yaml
spring:
  devtools:
    restart:
      trigger-file: ".reloadtrigger"

Restarts will now only happen when the src/main/resources/.reloadtrigger is updated.

You might want to set spring.devtools.restart.trigger-file as a global setting, so that all your projects behave in the same way.

Some IDEs have features that save you from needing to update your trigger file manually. Spring Tools for Eclipse and IntelliJ IDEA (Ultimate Edition) both have such support. With Spring Tools, you can use the “reload” button from the console view (as long as your trigger-file is named .reloadtrigger). For IntelliJ IDEA, you can follow the instructions in their documentation.

8.3.6. Customizing the Restart Classloader

As described earlier in the Restart vs Reload section, restart functionality is implemented by using two classloaders. If this causes issues, you can diagnose the problem by using the spring.devtools.restart.enabled system property, and if the app works with restart switched off, you might need to customize what gets loaded by which classloader.

By default, any open project in your IDE is loaded with the “restart” classloader, and any regular .jar file is loaded with the “base” classloader. The same is true if you use mvn spring-boot:run or gradle bootRun: the project containing your @SpringBootApplication is loaded with the “restart” classloader, and everything else with the “base” classloader. The classpath is printed on the console when you start the app, which can help to identify any problematic entries. Classes used reflectively, especially annotations, can be loaded into the parent (fixed) classloader on startup before the application classes which uses them, and this might lead to them not being detected by Spring in the application.

You can instruct Spring Boot to load parts of your project with a different classloader by creating a META-INF/spring-devtools.properties file. The spring-devtools.properties file can contain properties prefixed with restart.exclude and restart.include. The include elements are items that should be pulled up into the “restart” classloader, and the exclude elements are items that should be pushed down into the “base” classloader. The value of the property is a regex pattern that is applied to the classpath passed to the JVM on startup. Here is an example where some local class files are excluded and some extra libraries are included in the restart class loader:

Properties
restart.exclude.companycommonlibs=/mycorp-common-[\\w\\d-\\.]/(build|bin|out|target)/
restart.include.projectcommon=/mycorp-myproj-[\\w\\d-\\.]+\\.jar
Yaml
restart:
  exclude:
    companycommonlibs: "/mycorp-common-[\\w\\d-\\.]/(build|bin|out|target)/"
  include:
    projectcommon: "/mycorp-myproj-[\\w\\d-\\.]+\\.jar"
All property keys must be unique. As long as a property starts with restart.include. or restart.exclude. it is considered.
All META-INF/spring-devtools.properties from the classpath are loaded. You can package files inside your project, or in the libraries that the project consumes. System properties can not be used, only the properties file.

8.3.7. Known Limitations

Restart functionality does not work well with objects that are deserialized by using a standard ObjectInputStream. If you need to deserialize data, you may need to use Spring’s ConfigurableObjectInputStream in combination with Thread.currentThread().getContextClassLoader().

Unfortunately, several third-party libraries deserialize without considering the context classloader. If you find such a problem, you need to request a fix with the original authors.

8.4. LiveReload

The spring-boot-devtools module includes an embedded LiveReload server that can be used to trigger a browser refresh when a resource is changed. LiveReload browser extensions are freely available for Chrome, Firefox and Safari. You can find these extensions by searching 'LiveReload' in the marketplace or store of your chosen browser.

If you do not want to start the LiveReload server when your application runs, you can set the spring.devtools.livereload.enabled property to false.

You can only run one LiveReload server at a time. Before starting your application, ensure that no other LiveReload servers are running. If you start multiple applications from your IDE, only the first has LiveReload support.
To trigger LiveReload when a file changes, Automatic Restart must be enabled.

8.5. Global Settings

You can configure global devtools settings by adding any of the following files to the $HOME/.config/spring-boot directory:

  1. spring-boot-devtools.properties

  2. spring-boot-devtools.yaml

  3. spring-boot-devtools.yml

Any properties added to these files apply to all Spring Boot applications on your machine that use devtools. For example, to configure restart to always use a trigger file, you would add the following property to your spring-boot-devtools file:

Properties
spring.devtools.restart.trigger-file=.reloadtrigger
Yaml
spring:
  devtools:
    restart:
      trigger-file: ".reloadtrigger"

By default, $HOME is the user’s home directory. To customize this location, set the SPRING_DEVTOOLS_HOME environment variable or the spring.devtools.home system property.

If devtools configuration files are not found in $HOME/.config/spring-boot, the root of the $HOME directory is searched for the presence of a .spring-boot-devtools.properties file. This allows you to share the devtools global configuration with applications that are on an older version of Spring Boot that does not support the $HOME/.config/spring-boot location.

Profiles are not supported in devtools properties/yaml files.

Any profiles activated in .spring-boot-devtools.properties will not affect the loading of profile-specific configuration files. Profile specific filenames (of the form spring-boot-devtools-<profile>.properties) and spring.config.activate.on-profile documents in both YAML and Properties files are not supported.

8.5.1. Configuring File System Watcher

FileSystemWatcher works by polling the class changes with a certain time interval, and then waiting for a predefined quiet period to make sure there are no more changes. Since Spring Boot relies entirely on the IDE to compile and copy files into the location from where Spring Boot can read them, you might find that there are times when certain changes are not reflected when devtools restarts the application. If you observe such problems constantly, try increasing the spring.devtools.restart.poll-interval and spring.devtools.restart.quiet-period parameters to the values that fit your development environment:

Properties
spring.devtools.restart.poll-interval=2s
spring.devtools.restart.quiet-period=1s
Yaml
spring:
  devtools:
    restart:
      poll-interval: "2s"
      quiet-period: "1s"

The monitored classpath directories are now polled every 2 seconds for changes, and a 1 second quiet period is maintained to make sure there are no additional class changes.

8.6. Remote Applications

The Spring Boot developer tools are not limited to local development. You can also use several features when running applications remotely. Remote support is opt-in as enabling it can be a security risk. It should only be enabled when running on a trusted network or when secured with SSL. If neither of these options is available to you, you should not use DevTools' remote support. You should never enable support on a production deployment.

To enable it, you need to make sure that devtools is included in the repackaged archive, as shown in the following listing:

<build>
    <plugins>
        <plugin>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-maven-plugin</artifactId>
            <configuration>
                <excludeDevtools>false</excludeDevtools>
            </configuration>
        </plugin>
    </plugins>
</build>

Then you need to set the spring.devtools.remote.secret property. Like any important password or secret, the value should be unique and strong such that it cannot be guessed or brute-forced.

Remote devtools support is provided in two parts: a server-side endpoint that accepts connections and a client application that you run in your IDE. The server component is automatically enabled when the spring.devtools.remote.secret property is set. The client component must be launched manually.

Remote devtools is not supported for Spring WebFlux applications.

8.6.1. Running the Remote Client Application

The remote client application is designed to be run from within your IDE. You need to run org.springframework.boot.devtools.RemoteSpringApplication with the same classpath as the remote project that you connect to. The application’s single required argument is the remote URL to which it connects.

For example, if you are using Eclipse or Spring Tools and you have a project named my-app that you have deployed to Cloud Foundry, you would do the following:

  • Select Run Configurations…​ from the Run menu.

  • Create a new Java Application “launch configuration”.

  • Browse for the my-app project.

  • Use org.springframework.boot.devtools.RemoteSpringApplication as the main class.

  • Add https://myapp.cfapps.io to the Program arguments (or whatever your remote URL is).

A running remote client might resemble the following listing:

  .   ____          _                                              __ _ _
 /\\ / ___'_ __ _ _(_)_ __  __ _          ___               _      \ \ \ \
( ( )\___ | '_ | '_| | '_ \/ _` |        | _ \___ _ __  ___| |_ ___ \ \ \ \
 \\/  ___)| |_)| | | | | || (_| []::::::[]   / -_) '  \/ _ \  _/ -_) ) ) ) )
  '  |____| .__|_| |_|_| |_\__, |        |_|_\___|_|_|_\___/\__\___|/ / / /
 =========|_|==============|___/===================================/_/_/_/
 :: Spring Boot Remote ::  (v3.2.10)

2024-09-19T06:16:25.517Z  INFO 108272 --- [           main] o.s.b.devtools.RemoteSpringApplication   : Starting RemoteSpringApplication v3.2.10 using Java 17.0.12 with PID 108272 (/Users/myuser/.m2/repository/org/springframework/boot/spring-boot-devtools/3.2.10/spring-boot-devtools-3.2.10.jar started by myuser in /opt/apps/)
2024-09-19T06:16:25.522Z  INFO 108272 --- [           main] o.s.b.devtools.RemoteSpringApplication   : No active profile set, falling back to 1 default profile: "default"
2024-09-19T06:16:26.210Z  INFO 108272 --- [           main] o.s.b.d.a.OptionalLiveReloadServer       : LiveReload server is running on port 35729
2024-09-19T06:16:26.309Z  INFO 108272 --- [           main] o.s.b.devtools.RemoteSpringApplication   : Started RemoteSpringApplication in 2.099 seconds (process running for 3.411)
Because the remote client is using the same classpath as the real application it can directly read application properties. This is how the spring.devtools.remote.secret property is read and passed to the server for authentication.
It is always advisable to use https:// as the connection protocol, so that traffic is encrypted and passwords cannot be intercepted.
If you need to use a proxy to access the remote application, configure the spring.devtools.remote.proxy.host and spring.devtools.remote.proxy.port properties.

8.6.2. Remote Update

The remote client monitors your application classpath for changes in the same way as the local restart. Any updated resource is pushed to the remote application and (if required) triggers a restart. This can be helpful if you iterate on a feature that uses a cloud service that you do not have locally. Generally, remote updates and restarts are much quicker than a full rebuild and deploy cycle.

On a slower development environment, it may happen that the quiet period is not enough, and the changes in the classes may be split into batches. The server is restarted after the first batch of class changes is uploaded. The next batch can’t be sent to the application, since the server is restarting.

This is typically manifested by a warning in the RemoteSpringApplication logs about failing to upload some of the classes, and a consequent retry. But it may also lead to application code inconsistency and failure to restart after the first batch of changes is uploaded. If you observe such problems constantly, try increasing the spring.devtools.restart.poll-interval and spring.devtools.restart.quiet-period parameters to the values that fit your development environment. See the Configuring File System Watcher section for configuring these properties.

Files are only monitored when the remote client is running. If you change a file before starting the remote client, it is not pushed to the remote server.

9. Packaging Your Application for Production

Executable jars can be used for production deployment. As they are self-contained, they are also ideally suited for cloud-based deployment.

For additional “production ready” features, such as health, auditing, and metric REST or JMX end-points, consider adding spring-boot-actuator. See actuator.html for details.

10. What to Read Next

You should now understand how you can use Spring Boot and some best practices that you should follow. You can now go on to learn about specific Spring Boot features in depth, or you could skip ahead and read about the “production ready” aspects of Spring Boot.