I’m not going to explain what reactive programming is or why you should use it. I hope you’ve already read about it somewhere, and if not, you can Google it. In this post, I’m going to tell you how to use reactive programming specifically with Spring Boot and RxJava. Let's get started.

Prerequisites

Before you continue reading, I expect you understand how to create simple REST API using Spring Boot and RxJava. If you haven’t, you can learn more about Spring Boot on Baeldung and you can learn more about RxJava on AndroidHive. They explain those two materials really well.

Reactive REST API

The reactive REST API to be built is just a simple CRUD with authors and books. Here are the endpoints:

 [POST] /api/authors  → add an author

 [POST] /api/books  → add a book

 [PUT] /api/books/{bookId}  → update a book

 [GET] /api/books?limit={limit}&page={page}  → get list of books

 [GET] /api/book/{bookId}  → get a book’s detail

 [DELETE] /api/book/{bookId}  → delete a book

Dependencies

Open your pom.xml and add these dependencies.

<dependencies>
        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-web</artifactId>
            <version>2.1.5.RELEASE</version>
        </dependency>
        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-data-jpa</artifactId>
            <version>2.1.5.RELEASE</version>
        </dependency>

        <dependency>
            <groupId>io.reactivex</groupId>
            <artifactId>rxjava</artifactId>
            <version>1.3.8</version>
        </dependency>
    <!--IMPORTANT!!! ADD THIS DEPENDENCY TO SOLVE HttpMediaNotAcceptableException-->
        <dependency>
            <groupId>io.reactivex</groupId>
            <artifactId>rxjava-reactive-streams</artifactId>
            <version>1.2.1</version>
        </dependency>
        <dependency>
            <groupId>com.h2database</groupId>
            <artifactId>h2</artifactId>
            <version>1.4.199</version>
        </dependency>
        <dependency>
            <groupId>org.projectlombok</groupId>
            <artifactId>lombok</artifactId>
            <optional>true</optional>
            <version>1.18.8</version>
        </dependency>

        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-test</artifactId>
            <version>2.1.5.RELEASE</version>
            <scope>test</scope>
        </dependency>
        <dependency>
            <groupId>org.mockito</groupId>
            <artifactId>mockito-core</artifactId>
            <version>2.25.0</version>
            <scope>test</scope>
        </dependency>
    </dependencies>

P.S. Keep in mind that you have to add the dependency in line 19–23. If you don’t add it as a dependency, you’ll get HttpMediaNotAcceptableException every time you hit the reactive API. As you can see, I also added mockito as a dependency for mocking objects in unit tests. But I’ll cover the unit testing in another article.

Repository

The repository layer is just a regular JPA repository. Below is an example of the BookRepository.

@Repository
public interface BookRepository extends JpaRepository<Book, String> {
    List<Book> findAllByAuthorId(String authorId);
}

Service

For the service layer, the return value is not just regular data types, but I wrapped them inside RxJava’s Single. For example, the code below handles the addition of a new book.

 @Override
    public Single<String> addBook(AddBookRequest addBookRequest) {
        return saveBookToRepository(addBookRequest);
    }

    private Single<String> saveBookToRepository(AddBookRequest addBookRequest) {
        return Single.create(singleSubscriber -> {
            Optional<Author> optionalAuthor = authorRepository.findById(addBookRequest.getAuthorId());
            if (!optionalAuthor.isPresent())
                singleSubscriber.onError(new EntityNotFoundException());
            else {
                String addedBookId = bookRepository.save(toBook(addBookRequest)).getId();
                singleSubscriber.onSuccess(addedBookId);
            }
        });
    }

    private Book toBook(AddBookRequest addBookRequest) {
        Book book = new Book();
        BeanUtils.copyProperties(addBookRequest, book);
        book.setId(UUID.randomUUID().toString());
        book.setAuthor(Author.builder()
                .id(addBookRequest.getAuthorId())
                .build());
        return book;
    }

As you can see, the return value of the addBook method is a String wrapped inside RxJava’s Single.

Web/Controller

@PostMapping(
            consumes = MediaType.APPLICATION_JSON_VALUE,
            produces = MediaType.APPLICATION_JSON_VALUE
    )
    public Single<ResponseEntity<BaseWebResponse>> addBook(@RequestBody AddBookWebRequest addBookWebRequest) {
        return bookService.addBook(toAddBookRequest(addBookWebRequest))
                .subscribeOn(Schedulers.io())
                .map(s -> ResponseEntity.created(URI.create("/api/books/" + s)).body(BaseWebResponse.successNoData()));
    }

    private AddBookRequest toAddBookRequest(AddBookWebRequest addBookWebRequest) {
        AddBookRequest addBookRequest = new AddBookRequest();
        BeanUtils.copyProperties(addBookWebRequest, addBookRequest);
        return addBookRequest;
    }

In the web layer, it just forwards the request to the corresponding service, as shown above for handling the addition of a new book.

Just For Fun

Just for fun, I limited the threads down to 10 threads (default 200 threads) by adding this line of code inside application.properties.

server.tomcat.max-threads=10

Then, I did load testing using Apache Benchmark for the API (both blocking and non-blocking or reactive). I did 10k requests with 100 concurrency requests for the [POST] /api/books endpoint.

Blocking API

Non-Blocking/Reactive API

The End

The whole codes (+ unit tests) can be found on GitHub.