The Apache HTTP Client is a very robust library, suitable for both simple and advanced use cases when testing HTTP endpoints. Check out our guide covering basic request and response handling, as well as security, cookies, timeouts, and more:
Mocking is an essential part of unit testing, and the Mockito library makes it easy to write clean and intuitive unit tests for your Java code.
Get started with mocking and improve your application tests using our Mockito guide:
Handling concurrency in an application can be a tricky process with many potential pitfalls. A solid grasp of the fundamentals will go a long way to help minimize these issues.
Get started with understanding multi-threaded applications with our Java Concurrency guide:
Spring 5 added support for reactive programming with the Spring WebFlux module, which has been improved upon ever since. Get started with the Reactor project basics and reactive programming in Spring Boot:
Since its introduction in Java 8, the Stream API has become a staple of Java development. The basic operations like iterating, filtering, mapping sequences of elements are deceptively simple to use.
But these can also be overused and fall into some common pitfalls.
To get a better understanding on how Streams work and how to combine them with other language features, check out our guide to Java Streams:
Get started with Spring and Spring Boot, through the Learn Spring course:
>> LEARN SPRINGExplore Spring Boot 3 and Spring 6 in-depth through building a full REST API with the framework:
Yes, Spring Security can be complex, from the more advanced functionality within the Core to the deep OAuth support in the framework.
I built the security material as two full courses - Core and OAuth, to get practical with these more complex scenarios. We explore when and how to use each feature and code through it on the backing project.
You can explore the course here:
Spring Data JPA is a great way to handle the complexity of JPA with the powerful simplicity of Spring Boot.
Get started with Spring Data JPA through the guided reference course:
Refactor Java code safely β and automatically β with OpenRewrite.
Refactoring big codebases by hand is slow, risky, and easy to put off. Thatβs where OpenRewrite comes in. The open-source framework for large-scale, automated code transformations helps teams modernize safely and consistently.
Each month, the creators and maintainers of OpenRewrite at Moderne run live, hands-on training sessions β one for newcomers and one for experienced users. Youβll see how recipes work, how to apply them across projects, and how to modernize code with confidence.
Join the next session, bring your questions, and learn how to automate the kind of work that usually eats your sprint time.
1. Introduction
In this tutorial, weβll look at how we can retry HTTP requests when using Apache HtttpClient. Weβll also explore the default behavior in terms of the libraryβs retries and ways to configure it.
2. Default Retry Policy
Before we jump into the default behavior, weβll create a test class with the HttpClient instance and the requests counter:
public class ApacheHttpClientRetryUnitTest {
private Integer requestCounter;
private CloseableHttpClient httpClient;
@BeforeEach
void setUp() {
requestCounter = 0;
}
@AfterEach
void tearDown() throws IOException {
if (httpClient != null) {
httpClient.close();
}
}
}
Letβs start with the default behavior β Apache HttpClient retries at most 3 times all idempotent requests completed with IOException, so weβll get 4 requests in total. Weβll create here HttpClient that throws IOException for every request, just for the sake of demonstration:
private void createFailingHttpClient() {
this.httpClient = HttpClientBuilder
.create()
.addInterceptorFirst((HttpRequestInterceptor) (request, context) -> requestCounter++)
.addInterceptorLast((HttpResponseInterceptor) (response, context) -> { throw new IOException(); })
.build()
}
@Test
public void givenDefaultConfiguration_whenReceviedIOException_thenRetriesPerformed() {
createFailingHttpClient();
assertThrows(IOException.class, () -> httpClient.execute(new HttpGet("https://httpstat.us/200")));
assertThat(requestCounter).isEqualTo(4);
}
There are some IOException subclasses that HttpClient considers non-retryable. More specifically, they are:
- InterruptedIOException
- ConnectException
- UnknownHostException
- SSLException
- NoRouteToHostException
For instance, if we are unable to resolve a DNS name of the target host, then the request wonβt be retried:
public void createDefaultApacheHttpClient() {
this.httpClient = HttpClientBuilder
.create()
.addInterceptorFirst((HttpRequestInterceptor) (httpRequest, httpContext) -> {
requestCounter++;
}).build();
}
@Test
public void givenDefaultConfiguration_whenDomainNameNotResolved_thenNoRetryApplied() {
createDefaultApacheHttpClient();
HttpGet request = new HttpGet(URI.create("http://domain.that.does.not.exist:80/api/v1"));
assertThrows(UnknownHostException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(1);
}
As we can notice, these exceptions typically indicate a network or TLS problems. Therefore, they do not relate to unsuccessful HTTP request processing. It means that if the server has responded to our request with 5xx or 4xx, then there will be no retry logic applied:
@Test
public void givenDefaultConfiguration_whenGotInternalServerError_thenNoRetryLogicApplied() throws IOException {
createDefaultApacheHttpClient();
HttpGet request = new HttpGet(URI.create("https://httpstat.us/500"));
CloseableHttpResponse response = assertDoesNotThrow(() -> httpClient.execute(request));
assertThat(response.getStatusLine().getStatusCode()).isEqualTo(500);
assertThat(requestCounter).isEqualTo(1);
response.close();
}
But in most cases, that is not what we want. We typically want to retry on 5xx status codes at least. So, we will need to override the default behavior. Weβll do it in the next section.
3. Idempotency
Before our retries customization, we need to elaborate a bit on the idempotency of requests. It is important since the Apache HTTP client considers all HttpEntityEnclosingRequest implementations non-idempotent. Common implementations of this interface are HttpPost, HttpPut, and HttpPatch classes. So, our PATCH and PUT requests will not be, by default, retried:
@Test
public void givenDefaultConfiguration_whenHttpPatchRequest_thenRetryIsNotApplied() {
createFailingHttpClient();
HttpPatch request = new HttpPatch(URI.create("https://httpstat.us/500"));
assertThrows(IOException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(1);
}
@Test
public void givenDefaultConfiguration_whenHttpPutRequest_thenRetryIsNotApplied() {
createFailingHttpClient();
HttpPut request = new HttpPut(URI.create("https://httpstat.us/500"));
assertThrows(IOException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(1);
}
As we can see, there were no retries performed. Even if we received an IOException.
4. Custom RetryHandler
The default behavior we mentioned can be overridden. First, we can set the RetryHandler. For this, there is an option to use DefaultHttpRequestRetryHandler. This is a convenient out-of-the-box implementation of RetryHandler, which by the way, the library uses by default. This default implementation also implements the default behavior we discussed.
By using DefaultHttpRequestRetryHandler, we can set the number of retries we want and whenever HttpClient should retry idempotent requests or not:
private void createHttpClientWithRetryHandler() {
this.httpClient = HttpClientBuilder
.create()
.addInterceptorFirst((HttpRequestInterceptor) (httpRequest, httpContext) -> requestCounter++)
.addInterceptorLast((HttpResponseInterceptor) (httpRequest, httpContext) -> { throw new IOException(); })
.setRetryHandler(new DefaultHttpRequestRetryHandler(6, true))
.build();
}
@Test
public void givenConfiguredRetryHandler_whenHttpPostRequest_thenRetriesPerformed() {
createHttpClientWithRetryHandler();
HttpPost request = new HttpPost(URI.create("https://httpstat.us/500"));
assertThrows(IOException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(7);
}
As we can see, we configured the DefaultHttpRequestRetryHandler to make 6 retries. See the first constructor parameter. Also, we enabled the retry of idempotent requests. See the second constructor boolean parameter. Therefore, HttpCleint executes our POST request 7 times β 1 original request and 6 retries.
Also, if this level of customization is not enough, we can create our own RetryHandler:
private void createHttpClientWithCustomRetryHandler() {
this.httpClient = HttpClientBuilder
.create()
.addInterceptorFirst((HttpRequestInterceptor) (httpRequest, httpContext) -> requestCounter++)
.addInterceptorLast((HttpResponseInterceptor) (httpRequest, httpContext) -> { throw new IOException(); })
.setRetryHandler((exception, executionCount, context) -> {
if (executionCount <= 4 && Objects.equals("GET", ((HttpClientContext) context).getRequest().getRequestLine().getMethod())) {
return true;
} else {
return false;
}
}).build();
}
@Test
public void givenCustomRetryHandler_whenUnknownHostException_thenRetryAnyway() {
createHttpClientWithCustomRetryHandler();
HttpGet request = new HttpGet(URI.create("https://domain.that.does.not.exist/200"));
assertThrows(IOException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(5);
}
Here we basically said β to retry all GET requests 4 times regardless of the exception that occurred. So in the example above, we have retried UnknownHostException.
5. Disabling Retry Logic
Finally, there are cases when we would want to disable the reties. We can provide a RetryHandler that just always returns false, or we can use disableAutomaticRetries():
private void createHttpClientWithRetriesDisabled() {
this.httpClient = HttpClientBuilder
.create()
.addInterceptorFirst((HttpRequestInterceptor) (httpRequest, httpContext) -> requestCounter++)
.addInterceptorLast((HttpResponseInterceptor) (httpRequest, httpContext) -> { throw new IOException(); })
.disableAutomaticRetries()
.build();
}
@Test
public void givenDisabledRetries_whenExecutedHttpRequestEndUpWithIOException_thenRetryIsNotApplied() {
createHttpClientWithRetriesDisabled();
HttpGet request = new HttpGet(URI.create("https://httpstat.us/200"));
assertThrows(IOException.class, () -> httpClient.execute(request));
assertThat(requestCounter).isEqualTo(1);
}
By calling disableAutomaticRetries() on the HttpClientBuilder, we disabled all retries in the HttpClient. That means that no requests will be retired.
6. Conclusion
In this tutorial, we have discussed the default retry behavior in Apache HttpClient. Out-of-the-box RetryHandler will retry idempotent requests 3 times, considering the exception occurred. However, we can configure the number of retries and the non-idempotent requests retry policy. Also, we can provide our own RetryHandler implementation for even further customization. Finally, we can disable the retires by calling a method on HttpClientBuilder during HttpClient construction.
