WebClient

Spring WebFlux includes a reactive, non-blocking WebClient for HTTP requests. The client has a functional, fluent API with reactive types for declarative composition, see web-reactive.adoc. WebFlux client and server rely on the same non-blocking codecs to encode and decode request and response content.

Internally WebClient delegates to an HTTP client library. By default, it uses Reactor Netty, there is built-in support for the Jetty reactive HttpClient, and others can be plugged in through a ClientHttpConnector.

Configuration

The simplest way to create a WebClient is through one of the static factory methods:

WebClient.create()
WebClient.create(String baseUrl)

The above methods use the Reactor Netty HttpClient with default settings and expect io.projectreactor.netty:reactor-netty to be on the classpath.

You can also use WebClient.builder() with further options:

uriBuilderFactory: Customized UriBuilderFactory to use as a base URL.
defaultHeader: Headers for every request.
defaultCookie: Cookies for every request.
defaultRequest: Consumer to customize every request.
filter: Client filter for every request.
exchangeStrategies: HTTP message reader/writer customizations.
clientConnector: HTTP client library settings.

The following example configures HTTP codecs:

	ExchangeStrategies strategies = ExchangeStrategies.builder()
			.codecs(configurer -> {
				// ...
			})
			.build();

	WebClient client = WebClient.builder()
			.exchangeStrategies(strategies)
			.build();

Once built, a WebClient instance is immutable. However, you can clone it and build a modified copy without affecting the original instance, as the following example shows:

	WebClient client1 = WebClient.builder()
			.filter(filterA).filter(filterB).build();

	WebClient client2 = client1.mutate()
			.filter(filterC).filter(filterD).build();

	// client1 has filterA, filterB

	// client2 has filterA, filterB, filterC, filterD

Reactor Netty

To customize Reactor Netty settings, simple provide a pre-configured HttpClient:

	HttpClient httpClient = HttpClient.create().secure(sslSpec -> ...);

	WebClient webClient = WebClient.builder()
			.clientConnector(new ReactorClientHttpConnector(httpClient))
			.build();

Resources

By default, HttpClient participates in the global Reactor Netty resources held in reactor.netty.http.HttpResources, including event loop threads and a connection pool. This is the recommended mode, since fixed, shared resources are preferred for event loop concurrency. In this mode global resources remain active until the process exits.

If the server is timed with the process, there is typically no need for an explicit shutdown. However, if the server can start or stop in-process (for example, a Spring MVC application deployed as a WAR), you can declare a Spring-managed bean of type ReactorResourceFactory with globalResources=true (the default) to ensure that the Reactor Netty global resources are shut down when the Spring ApplicationContext is closed, as the following example shows:

	@Bean
	public ReactorResourceFactory reactorResourceFactory() {
		return new ReactorResourceFactory();
	}

You can also choose not to participate in the global Reactor Netty resources. However, in this mode, the burden is on you to ensure that all Reactor Netty client and server instances use shared resources, as the following example shows:

	@Bean
	public ReactorResourceFactory resourceFactory() {
		ReactorResourceFactory factory = new ReactorResourceFactory();
		factory.setGlobalResources(false); (1)
		return factory;
	}

	@Bean
	public WebClient webClient() {

		Function<HttpClient, HttpClient> mapper = client -> {
			// Further customizations...
		};

		ClientHttpConnector connector =
				new ReactorClientHttpConnector(resourceFactory(), mapper); (2)

		return WebClient.builder().clientConnector(connector).build(); (3)
	}

Create resources independent of global ones.
Use the ReactorClientHttpConnector constructor with resource factory.
Plug the connector into the WebClient.Builder.

Timeouts

To configure a connection timeout:

import io.netty.channel.ChannelOption;

HttpClient httpClient = HttpClient.create()
		.tcpConfiguration(client ->
				client.option(ChannelOption.CONNECT_TIMEOUT_MILLIS, 10000));

To configure a read and/or write timeout values:

import io.netty.handler.timeout.ReadTimeoutHandler;
import io.netty.handler.timeout.WriteTimeoutHandler;

HttpClient httpClient = HttpClient.create()
		.tcpConfiguration(client ->
				client.doOnConnected(conn -> conn
						.addHandlerLast(new ReadTimeoutHandler(10))
						.addHandlerLast(new WriteTimeoutHandler(10))));

Jetty

The following example shows how to customize Jetty HttpClient settings:

	HttpClient httpClient = new HttpClient();
	httpClient.setCookieStore(...);
	ClientHttpConnector connector = new JettyClientHttpConnector(httpClient);

	WebClient webClient = WebClient.builder().clientConnector(connector).build();

By default, HttpClient creates its own resources (Executor, ByteBufferPool, Scheduler), which remain active until the process exits or stop() is called.

You can share resources between multiple instances of the Jetty client (and server) and ensure that the resources are shut down when the Spring ApplicationContext is closed by declaring a Spring-managed bean of type JettyResourceFactory, as the following example shows:

	@Bean
	public JettyResourceFactory resourceFactory() {
		return new JettyResourceFactory();
	}

	@Bean
	public WebClient webClient() {

		Consumer<HttpClient> customizer = client -> {
			// Further customizations...
		};

		ClientHttpConnector connector =
				new JettyClientHttpConnector(resourceFactory(), customizer); (1)

		return WebClient.builder().clientConnector(connector).build(); (2)
	}

Use the JettyClientHttpConnector constructor with resource factory.
Plug the connector into the WebClient.Builder.

`retrieve()`

The retrieve() method is the easiest way to get a response body and decode it. The following example shows how to do so:

	WebClient client = WebClient.create("https://example.org");

	Mono<Person> result = client.get()
			.uri("/persons/{id}", id).accept(MediaType.APPLICATION_JSON)
			.retrieve()
			.bodyToMono(Person.class);

You can also get a stream of objects decoded from the response, as the following example shows:

	Flux<Quote> result = client.get()
			.uri("/quotes").accept(MediaType.TEXT_EVENT_STREAM)
			.retrieve()
			.bodyToFlux(Quote.class);

By default, responses with 4xx or 5xx status codes result in an WebClientResponseException or one of its HTTP status specific sub-classes, such as WebClientResponseException.BadRequest, WebClientResponseException.NotFound, and others. You can also use the onStatus method to customize the resulting exception, as the following example shows:

	Mono<Person> result = client.get()
			.uri("/persons/{id}", id).accept(MediaType.APPLICATION_JSON)
			.retrieve()
			.onStatus(HttpStatus::is4xxClientError, response -> ...)
			.onStatus(HttpStatus::is5xxServerError, response -> ...)
			.bodyToMono(Person.class);

When onStatus is used, if the response is expected to have content, then the onStatus callback should consume it. If not, the content will be automatically drained to ensure resources are released.

`exchange()`

The exchange() method provides more control than the retrieve method. The following example is equivalent to retrieve() but also provides access to the ClientResponse:

	Mono<Person> result = client.get()
			.uri("/persons/{id}", id).accept(MediaType.APPLICATION_JSON)
			.exchange()
			.flatMap(response -> response.bodyToMono(Person.class));

At this level, you can also create a full ResponseEntity:

	Mono<ResponseEntity<Person>> result = client.get()
			.uri("/persons/{id}", id).accept(MediaType.APPLICATION_JSON)
			.exchange()
			.flatMap(response -> response.toEntity(Person.class));

Note that (unlike retrieve()), with exchange(), there are no automatic error signals for 4xx and 5xx responses. You have to check the status code and decide how to proceed.

Caution

When you use exchange(), you must always use any of the body or toEntity methods of ClientResponse to ensure resources are released and to avoid potential issues with HTTP connection pooling. You can use bodyToMono(Void.class) if no response content is expected. However, if the response does have content, the connection is closed and is not placed back in the pool.

Request Body

The request body can be encoded from an Object, as the following example shows:

	Mono<Person> personMono = ... ;

	Mono<Void> result = client.post()
			.uri("/persons/{id}", id)
			.contentType(MediaType.APPLICATION_JSON)
			.body(personMono, Person.class)
			.retrieve()
			.bodyToMono(Void.class);

You can also have a stream of objects be encoded, as the following example shows:

	Flux<Person> personFlux = ... ;

	Mono<Void> result = client.post()
			.uri("/persons/{id}", id)
			.contentType(MediaType.APPLICATION_STREAM_JSON)
			.body(personFlux, Person.class)
			.retrieve()
			.bodyToMono(Void.class);

Alternatively, if you have the actual value, you can use the syncBody shortcut method, as the following example shows:

	Person person = ... ;

	Mono<Void> result = client.post()
			.uri("/persons/{id}", id)
			.contentType(MediaType.APPLICATION_JSON)
			.syncBody(person)
			.retrieve()
			.bodyToMono(Void.class);

Form Data

To send form data, you can provide a MultiValueMap<String, String> as the body. Note that the content is automatically set to application/x-www-form-urlencoded by the FormHttpMessageWriter. The following example shows how to use MultiValueMap<String, String>:

	MultiValueMap<String, String> formData = ... ;

	Mono<Void> result = client.post()
			.uri("/path", id)
			.syncBody(formData)
			.retrieve()
			.bodyToMono(Void.class);

You can also supply form data in-line by using BodyInserters, as the following example shows:

	import static org.springframework.web.reactive.function.BodyInserters.*;

	Mono<Void> result = client.post()
			.uri("/path", id)
			.body(fromFormData("k1", "v1").with("k2", "v2"))
			.retrieve()
			.bodyToMono(Void.class);

Multipart Data

To send multipart data, you need to provide a MultiValueMap<String, ?> whose values are either Object instances that represent part content or HttpEntity instances that represent the content and headers for a part. MultipartBodyBuilder provides a convenient API to prepare a multipart request. The following example shows how to create a MultiValueMap<String, ?>:

	MultipartBodyBuilder builder = new MultipartBodyBuilder();
	builder.part("fieldPart", "fieldValue");
	builder.part("filePart", new FileSystemResource("...logo.png"));
	builder.part("jsonPart", new Person("Jason"));

	MultiValueMap<String, HttpEntity<?>> parts = builder.build();

In most cases, you do not have to specify the Content-Type for each part. The content type is determined automatically based on the HttpMessageWriter chosen to serialize it or, in the case of a Resource, based on the file extension. If necessary, you can explicitly provide the MediaType to use for each part through one of the overloaded builder part methods.

Once a MultiValueMap is prepared, the easiest way to pass it to the the WebClient is through the syncBody method, as the following example shows:

	MultipartBodyBuilder builder = ...;

	Mono<Void> result = client.post()
			.uri("/path", id)
			.syncBody(builder.build())
			.retrieve()
			.bodyToMono(Void.class);

If the MultiValueMap contains at least one non-String value, which could also represent regular form data (that is, application/x-www-form-urlencoded), you need not set the Content-Type to multipart/form-data. This is always the case when using MultipartBodyBuilder, which ensures an HttpEntity wrapper.

As an alternative to MultipartBodyBuilder, you can also provide multipart content, inline-style, through the built-in BodyInserters, as the following example shows:

	import static org.springframework.web.reactive.function.BodyInserters.*;

	Mono<Void> result = client.post()
			.uri("/path", id)
			.body(fromMultipartData("fieldPart", "value").with("filePart", resource))
			.retrieve()
			.bodyToMono(Void.class);

Client Filters

You can register a client filter (ExchangeFilterFunction) through the WebClient.Builder in order to intercept and modify requests, as the following example shows:

WebClient client = WebClient.builder()
		.filter((request, next) -> {

			ClientRequest filtered = ClientRequest.from(request)
					.header("foo", "bar")
					.build();

			return next.exchange(filtered);
		})
		.build();

This can be used for cross-cutting concerns, such as authentication. The following example uses a filter for basic authentication through a static factory method:

// static import of ExchangeFilterFunctions.basicAuthentication

WebClient client = WebClient.builder()
		.filter(basicAuthentication("user", "password"))
		.build();

Filters apply globally to every request. To change a filter’s behavior for a specific request, you can add request attributes to the ClientRequest that can then be accessed by all filters in the chain, as the following example shows:

WebClient client = WebClient.builder()
		.filter((request, next) -> {
			Optional<Object> usr = request.attribute("myAttribute");
			// ...
		})
		.build();

client.get().uri("https://example.org/")
		.attribute("myAttribute", "...")
		.retrieve()
		.bodyToMono(Void.class);

	}

You can also replicate an existing WebClient, insert new filters, or remove already registered filters. The following example, inserts a basic authentication filter at index 0:

// static import of ExchangeFilterFunctions.basicAuthentication

WebClient client = webClient.mutate()
		.filters(filterList -> {
			filterList.add(0, basicAuthentication("user", "password"));
		})
		.build();

Synchronous Use

WebClient can be used in synchronous style by blocking at the end for the result:

Person person = client.get().uri("/person/{id}", i).retrieve()
	.bodyToMono(Person.class)
	.block();

List<Person> persons = client.get().uri("/persons").retrieve()
	.bodyToFlux(Person.class)
	.collectList()
	.block();

However if multiple calls need to be made, it’s more efficient to avoid blocking on each response individually, and instead wait for the combined result:

Mono<Person> personMono = client.get().uri("/person/{id}", personId)
		.retrieve().bodyToMono(Person.class);

Mono<List<Hobby>> hobbiesMono = client.get().uri("/person/{id}/hobbies", personId)
		.retrieve().bodyToFlux(Hobby.class).collectList();

Map<String, Object> data = Mono.zip(personMono, hobbiesMono, (person, hobbies) -> {
			Map<String, String> map = new LinkedHashMap<>();
			map.put("person", personName);
			map.put("hobbies", hobbies);
			return map;
		})
		.block();

The above is merely one example. There are lots of other patterns and operators for putting together a reactive pipeline that makes many remote calls, potentially some nested, inter-dependent, without ever blocking until the end.

Note	You should never have to block in a Spring MVC controller. Simply return the resulting `Flux` or `Mono` from the controller method.

Testing

To test code that uses the WebClient, you can use a mock web server, such as the OkHttp MockWebServer. To see an example of its use, check out WebClientIntegrationTests in the Spring Framework test suite or the static-server sample in the OkHttp repository.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

webflux-webclient.adoc

webflux-webclient.adoc

WebClient

Configuration

Reactor Netty

Resources

Timeouts

Jetty

`retrieve()`

`exchange()`

Request Body

Form Data

Multipart Data

Client Filters

Synchronous Use

Testing

Files

webflux-webclient.adoc

Latest commit

History

webflux-webclient.adoc

File metadata and controls

WebClient

Configuration

Reactor Netty

Resources

Timeouts

Jetty

retrieve()

exchange()

Request Body

Form Data

Multipart Data

Client Filters

Synchronous Use

Testing

`retrieve()`

`exchange()`