Spring Boot 3.x RestTemplate迁移至WebClient问题详解与解决方案

优质文章学习记录

— 19 min read

目录

Spring Boot 3.x RestTemplate迁移至WebClient问题详解与解决方案

一、RestTemplate弃用背景与迁移策略

1.1 弃用原因分析

// RestTemplate的典型用法（已弃用）@Component@Deprecated(since ="Spring 5.0", forRemoval =true)publicclassOldRestTemplateService{@AutowiredprivateRestTemplate restTemplate;publicUsergetUser(String userId){// ❌ 同步阻塞调用ResponseEntity<User> response = restTemplate.exchange("https://api.example.com/users/{id}",HttpMethod.GET,null,User.class, userId );return response.getBody();}publicList<User>getUsers(){// ❌ 类型擦除问题ParameterizedTypeReference<List<User>> typeRef =newParameterizedTypeReference<List<User>>(){};ResponseEntity<List<User>> response = restTemplate.exchange("https://api.example.com/users",HttpMethod.GET,null, typeRef );return response.getBody();}}// WebClient的优势@ComponentpublicclassWebClientMigrationGuide{// ✅ 非阻塞、响应式// ✅ 函数式API// ✅ 更好的错误处理// ✅ 流式处理支持// ✅ 背压支持}

1.2 迁移依赖配置

<!-- pom.xml - WebClient依赖 --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-webflux</artifactId></dependency><!-- 可选：Reactive MongoDB（如果需要） --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-data-mongodb-reactive</artifactId></dependency><!-- 可选：Reactive Redis --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-data-redis-reactive</artifactId></dependency><!-- 测试依赖 --><dependency><groupId>io.projectreactor</groupId><artifactId>reactor-test</artifactId><scope>test</scope></dependency>
# application.yml - WebClient配置spring:webflux:client:max-memory-size: 10MB codecs:default-codecs:max-in-memory-size: 10MB # HTTP客户端配置httpclient:connect-timeout: 5s response-timeout: 30s read-timeout: 30s write-timeout: 30s pool:max-connections:1000max-life-time: 30s acquire-timeout: 45s # 日志配置logging:level:reactor.netty.http.client: DEBUG org.springframework.web.reactive.function.client: DEBUG

二、核心迁移方案：RestTemplate ↔ WebClient映射

2.1 基础迁移对照表

@Component@Slf4jpublicclassMigrationMappingGuide{// ==================== 配置对比 ====================publicvoidconfigurationComparison(){// RestTemplate配置@Deprecated@BeanpublicRestTemplaterestTemplate(){RestTemplate restTemplate =newRestTemplate(); restTemplate.setMessageConverters(Arrays.asList(newMappingJackson2HttpMessageConverter(),newStringHttpMessageConverter())); restTemplate.setErrorHandler(newDefaultResponseErrorHandler()); restTemplate.setInterceptors(List.of(newLoggingInterceptor()));return restTemplate;}// WebClient配置（等价实现）@BeanpublicWebClientwebClient(){returnWebClient.builder().codecs(configurer ->{ configurer.defaultCodecs().maxInMemorySize(10*1024*1024);}).filter(ExchangeFilterFunction.ofRequestProcessor(request ->{ log.debug("Request: {} {}", request.method(), request.url());returnMono.just(request);})).filter(ExchangeFilterFunction.ofResponseProcessor(response ->{ log.debug("Response status: {}", response.statusCode());returnMono.just(response);})).build();}}// ==================== GET请求迁移 ====================publicvoidgetRequestMigration(){// RestTemplate GET@DeprecatedpublicUsergetById_RestTemplate(String id){ResponseEntity<User> response = restTemplate.getForEntity("https://api.example.com/users/{id}",User.class, id );return response.getBody();}// WebClient GET（等价实现）publicMono<User>getById_WebClient(String id){return webClient.get().uri("https://api.example.com/users/{id}", id).retrieve().bodyToMono(User.class);}// 带查询参数的GET@DeprecatedpublicList<User>search_RestTemplate(String name,int age){Map<String,Object> params =newHashMap<>(); params.put("name", name); params.put("age", age);ResponseEntity<List<User>> response = restTemplate.exchange("https://api.example.com/users?name={name}&age={age}",HttpMethod.GET,null,newParameterizedTypeReference<List<User>>(){}, params );return response.getBody();}publicFlux<User>search_WebClient(String name,int age){return webClient.get().uri(uriBuilder -> uriBuilder .path("/users").queryParam("name", name).queryParam("age", age).build()).retrieve().bodyToFlux(User.class);}}// ==================== POST请求迁移 ====================publicvoidpostRequestMigration(){// RestTemplate POST@DeprecatedpublicUsercreate_RestTemplate(User user){ResponseEntity<User> response = restTemplate.postForEntity("https://api.example.com/users", user,User.class);return response.getBody();}// WebClient POST（等价实现）publicMono<User>create_WebClient(User user){return webClient.post().uri("https://api.example.com/users").contentType(MediaType.APPLICATION_JSON).bodyValue(user).retrieve().bodyToMono(User.class);}// POST with headers@DeprecatedpublicUsercreateWithAuth_RestTemplate(User user,String token){HttpHeaders headers =newHttpHeaders(); headers.set("Authorization","Bearer "+ token); headers.setContentType(MediaType.APPLICATION_JSON);HttpEntity<User> request =newHttpEntity<>(user, headers);ResponseEntity<User> response = restTemplate.exchange("https://api.example.com/users",HttpMethod.POST, request,User.class);return response.getBody();}publicMono<User>createWithAuth_WebClient(User user,String token){return webClient.post().uri("https://api.example.com/users").header("Authorization","Bearer "+ token).contentType(MediaType.APPLICATION_JSON).bodyValue(user).retrieve().bodyToMono(User.class);}}// ==================== PUT/DELETE请求迁移 ====================publicvoidputDeleteMigration(){// RestTemplate PUT@Deprecatedpublicvoidupdate_RestTemplate(String id,User user){ restTemplate.put("https://api.example.com/users/{id}", user, id );}// WebClient PUTpublicMono<Void>update_WebClient(String id,User user){return webClient.put().uri("https://api.example.com/users/{id}", id).contentType(MediaType.APPLICATION_JSON).bodyValue(user).retrieve().bodyToMono(Void.class);}// RestTemplate DELETE@Deprecatedpublicvoiddelete_RestTemplate(String id){ restTemplate.delete("https://api.example.com/users/{id}", id );}// WebClient DELETEpublicMono<Void>delete_WebClient(String id){return webClient.delete().uri("https://api.example.com/users/{id}", id).retrieve().bodyToMono(Void.class);}}}

2.2 复杂场景迁移

@Component@Slf4jpublicclassComplexScenarioMigration{// ==================== 文件上传迁移 ====================publicvoidfileUploadMigration(){// RestTemplate 文件上传@DeprecatedpublicStringuploadFile_RestTemplate(MultipartFile file)throwsIOException{HttpHeaders headers =newHttpHeaders(); headers.setContentType(MediaType.MULTIPART_FORM_DATA);MultiValueMap<String,Object> body =newLinkedMultiValueMap<>(); body.add("file",newByteArrayResource(file.getBytes()){@OverridepublicStringgetFilename(){return file.getOriginalFilename();}});HttpEntity<MultiValueMap<String,Object>> request =newHttpEntity<>(body, headers);ResponseEntity<String> response = restTemplate.postForEntity("https://api.example.com/upload", request,String.class);return response.getBody();}// WebClient 文件上传publicMono<String>uploadFile_WebClient(MultipartFile file){returnMono.fromCallable(()-> file.getBytes()).flatMap(bytes ->{MultipartBodyBuilder builder =newMultipartBodyBuilder(); builder.part("file",newByteArrayResource(bytes){@OverridepublicStringgetFilename(){return file.getOriginalFilename();}});return webClient.post().uri("https://api.example.com/upload").contentType(MediaType.MULTIPART_FORM_DATA).body(BodyInserters.fromMultipartData(builder.build())).retrieve().bodyToMono(String.class);});}}// ==================== 流式响应迁移 ====================publicvoidstreamingResponseMigration(){// RestTemplate 处理流式响应（复杂）@DeprecatedpublicList<String>streamLines_RestTemplate(){ResponseEntity<Resource> response = restTemplate.exchange("https://api.example.com/stream",HttpMethod.GET,null,Resource.class);List<String> lines =newArrayList<>();try(BufferedReader reader =newBufferedReader(newInputStreamReader(response.getBody().getInputStream()))){String line;while((line = reader.readLine())!=null){ lines.add(line);}}catch(IOException e){thrownewRuntimeException(e);}return lines;}// WebClient 处理流式响应（简单）publicFlux<String>streamLines_WebClient(){return webClient.get().uri("https://api.example.com/stream").retrieve().bodyToFlux(String.class).doOnNext(line -> log.debug("Received line: {}", line));}}// ==================== 并发请求迁移 ====================publicvoidconcurrentRequestsMigration(){// RestTemplate 并发请求（使用CompletableFuture包装）@DeprecatedpublicCompletableFuture<List<User>>fetchUsersConcurrently_RestTemplate(List<String> userIds){List<CompletableFuture<User>> futures = userIds.stream().map(id ->CompletableFuture.supplyAsync(()-> restTemplate.getForObject("https://api.example.com/users/{id}",User.class, id))).collect(Collectors.toList());CompletableFuture<Void> allDone =CompletableFuture.allOf( futures.toArray(newCompletableFuture[0]));return allDone.thenApply(v -> futures.stream().map(CompletableFuture::join).collect(Collectors.toList()));}// WebClient 并发请求（响应式，更简洁）publicFlux<User>fetchUsersConcurrently_WebClient(List<String> userIds){returnFlux.fromIterable(userIds).flatMap(id -> webClient.get().uri("https://api.example.com/users/{id}", id).retrieve().bodyToMono(User.class),5);// 并发度为5}}// ==================== 请求拦截器迁移 ====================publicvoidinterceptorMigration(){// RestTemplate 拦截器@DeprecatedpublicclassRestTemplateAuthInterceptorimplementsClientHttpRequestInterceptor{@OverridepublicClientHttpResponseintercept(HttpRequest request,byte[] body,ClientHttpRequestExecution execution)throwsIOException{ request.getHeaders().add("Authorization","Bearer token");return execution.execute(request, body);}}// WebClient 过滤器（等效拦截器）publicExchangeFilterFunctionwebClientAuthFilter(){return(request, next)->{ClientRequest filteredRequest =ClientRequest.from(request).header("Authorization","Bearer token").build();return next.exchange(filteredRequest);};}// 添加过滤器到WebClient@BeanpublicWebClientfilteredWebClient(){returnWebClient.builder().filter(webClientAuthFilter()).filter(loggingFilter()).build();}privateExchangeFilterFunctionloggingFilter(){returnExchangeFilterFunction.ofRequestProcessor(request ->{ log.info("Request: {} {}", request.method(), request.url());returnMono.just(request);});}}}

三、错误处理机制迁移

3.1 错误处理对照

@Component@Slf4jpublicclassErrorHandlingMigration{// ==================== RestTemplate错误处理 ====================@DeprecatedpublicUsergetUserWithErrorHandling_RestTemplate(String userId){try{ResponseEntity<User> response = restTemplate.exchange("https://api.example.com/users/{id}",HttpMethod.GET,null,User.class, userId );return response.getBody();}catch(HttpClientErrorException e){if(e.getStatusCode()==HttpStatus.NOT_FOUND){thrownewUserNotFoundException("User not found: "+ userId);}elseif(e.getStatusCode()==HttpStatus.UNAUTHORIZED){thrownewAuthenticationException("Unauthorized access");}else{thrownewServiceException("Service error: "+ e.getMessage());}}catch(HttpServerErrorException e){thrownewServiceException("Server error: "+ e.getMessage());}catch(ResourceAccessException e){thrownewNetworkException("Network error: "+ e.getMessage());}}// ==================== WebClient错误处理 ====================publicMono<User>getUserWithErrorHandling_WebClient(String userId){return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().onStatus(status -> status ==HttpStatus.NOT_FOUND, response ->Mono.error(newUserNotFoundException("User not found: "+ userId))).onStatus(status -> status ==HttpStatus.UNAUTHORIZED, response ->Mono.error(newAuthenticationException("Unauthorized access"))).onStatus(HttpStatusCode::is5xxServerError, response ->Mono.error(newServiceException("Server error: "+ response.statusCode()))).bodyToMono(User.class).onErrorMap(WebClientRequestException.class, e ->newNetworkException("Network error: "+ e.getMessage())).doOnError(e -> log.error("Failed to get user: {}", userId, e));}// ==================== 重试机制迁移 ====================@DeprecatedpublicUsergetUserWithRetry_RestTemplate(String userId){// RestTemplate需要自定义重试逻辑int maxAttempts =3;int attempt =0;while(attempt < maxAttempts){try{ResponseEntity<User> response = restTemplate.exchange("https://api.example.com/users/{id}",HttpMethod.GET,null,User.class, userId );return response.getBody();}catch(ResourceAccessException e){ attempt++;if(attempt == maxAttempts){thrownewNetworkException("Max retries exceeded", e);}try{Thread.sleep(1000* attempt);// 简单退避}catch(InterruptedException ie){Thread.currentThread().interrupt();thrownewNetworkException("Interrupted", ie);}}}thrownewNetworkException("Failed after retries");}publicMono<User>getUserWithRetry_WebClient(String userId){// WebClient内置重试支持return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).retryWhen(Retry.backoff(3,Duration.ofSeconds(1)).maxBackoff(Duration.ofSeconds(10)).jitter(0.5).doBeforeRetry(retrySignal -> log.warn("Retrying user fetch, attempt {}", retrySignal.totalRetries()+1)).onRetryExhaustedThrow((retryBackoffSpec, retrySignal)->newNetworkException("Max retries exceeded")));}// ==================== 断路器模式迁移 ====================publicMono<User>getUserWithCircuitBreaker_WebClient(String userId){// 使用Resilience4j CircuitBreakerCircuitBreaker circuitBreaker =CircuitBreaker.ofDefaults("userService");returnMono.defer(()-> webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class)).transformDeferred(CircuitBreakerOperator.of(circuitBreaker)).doOnSuccess(user -> log.info("User fetched successfully: {}", userId)).doOnError(e -> log.error("Failed to fetch user: {}", userId, e));}// ==================== 超时处理迁移 ====================@DeprecatedpublicUsergetUserWithTimeout_RestTemplate(String userId){// RestTemplate需要配置ClientHttpRequestFactorySimpleClientHttpRequestFactory factory =newSimpleClientHttpRequestFactory(); factory.setConnectTimeout(5000); factory.setReadTimeout(10000); restTemplate.setRequestFactory(factory);try{return restTemplate.getForObject("https://api.example.com/users/{id}",User.class, userId);}catch(ResourceAccessException e){thrownewTimeoutException("Request timeout", e);}}publicMono<User>getUserWithTimeout_WebClient(String userId){// WebClient支持响应式超时return webClient.get().uri("https://api.example.com/users/{id}", userId).httpRequest(httpRequest ->{// 设置连接超时和读取超时HttpClientRequest reactorRequest = httpRequest.getNativeRequest(); reactorRequest.responseTimeout(Duration.ofSeconds(10));}).retrieve().bodyToMono(User.class).timeout(Duration.ofSeconds(10)).onErrorMap(TimeoutException.class, e ->newTimeoutException("Request timeout", e));}}

3.2 统一错误处理策略

@Component@Slf4jpublicclassUnifiedErrorHandlingStrategy{// 自定义异常类publicstaticclassApiExceptionextendsRuntimeException{privatefinalHttpStatus status;privatefinalString errorCode;publicApiException(HttpStatus status,String errorCode,String message){super(message);this.status = status;this.errorCode = errorCode;}}// WebClient错误处理器@BeanpublicWebClientwebClientWithGlobalErrorHandler(WebClient.Builder builder){return builder.filter(globalErrorHandler()).build();}privateExchangeFilterFunctionglobalErrorHandler(){returnExchangeFilterFunction.ofResponseProcessor(response ->{if(response.statusCode().isError()){return response.bodyToMono(String.class).defaultIfEmpty("").flatMap(body ->{HttpStatusCode status = response.statusCode();String message = body.isEmpty()? status.getReasonPhrase(): body;// 根据状态码创建特定异常returnMono.error(createApiException(status, message));});}returnMono.just(response);});}privateApiExceptioncreateApiException(HttpStatusCode status,String message){returnswitch(status.value()){case400->newApiException(HttpStatus.BAD_REQUEST,"BAD_REQUEST", message);case401->newApiException(HttpStatus.UNAUTHORIZED,"UNAUTHORIZED", message);case403->newApiException(HttpStatus.FORBIDDEN,"FORBIDDEN", message);case404->newApiException(HttpStatus.NOT_FOUND,"NOT_FOUND", message);case429->newApiException(HttpStatus.TOO_MANY_REQUESTS,"RATE_LIMITED", message);case500->newApiException(HttpStatus.INTERNAL_SERVER_ERROR,"SERVER_ERROR", message);case502->newApiException(HttpStatus.BAD_GATEWAY,"BAD_GATEWAY", message);case503->newApiException(HttpStatus.SERVICE_UNAVAILABLE,"SERVICE_UNAVAILABLE", message);case504->newApiException(HttpStatus.GATEWAY_TIMEOUT,"GATEWAY_TIMEOUT", message);default->newApiException(HttpStatus.valueOf(status.value()),"UNKNOWN_ERROR", message);};}// 全局异常处理器（Controller层）@ControllerAdvicepublicstaticclassGlobalExceptionHandler{@ExceptionHandler(ApiException.class)publicResponseEntity<ErrorResponse>handleApiException(ApiException ex){ErrorResponse error =newErrorResponse( ex.getErrorCode(), ex.getMessage(),Instant.now());returnnewResponseEntity<>(error, ex.getStatus());}@ExceptionHandler(TimeoutException.class)publicResponseEntity<ErrorResponse>handleTimeout(TimeoutException ex){ErrorResponse error =newErrorResponse("TIMEOUT","Request timeout",Instant.now());returnnewResponseEntity<>(error,HttpStatus.GATEWAY_TIMEOUT);}@Data@AllArgsConstructorpublicstaticclassErrorResponse{privateString code;privateString message;privateInstant timestamp;}}}

四、异步到同步的桥接方案

4.1 阻塞式适配器

@Component@Slf4jpublicclassBlockingAdapter{privatefinalWebClient webClient;privatefinalScheduler blockingScheduler;publicBlockingAdapter(WebClient.Builder webClientBuilder){this.webClient = webClientBuilder.build();this.blockingScheduler =Schedulers.boundedElastic();}// ==================== 简单阻塞调用 ====================@DeprecatedpublicUsergetUserBlocking_RestTemplate(String userId){return restTemplate.getForObject("https://api.example.com/users/{id}",User.class, userId);}// WebClient阻塞调用（不推荐，但可做迁移过渡）publicUsergetUserBlocking_WebClient(String userId){return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).block();// ⚠️ 阻塞调用，慎用}// ==================== 推荐的异步转同步模式 ====================publicUsergetUserWithTimeout(String userId){returngetUserAsync(userId).block(Duration.ofSeconds(10));// 指定超时时间}publicMono<User>getUserAsync(String userId){return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class);}// ==================== 批量阻塞调用 ====================@DeprecatedpublicList<User>getUsersBlocking_RestTemplate(List<String> userIds){List<User> users =newArrayList<>();for(String userId : userIds){User user = restTemplate.getForObject("https://api.example.com/users/{id}",User.class, userId); users.add(user);}return users;}// WebClient批量调用（推荐）publicList<User>getUsersBlocking_WebClient(List<String> userIds){returnFlux.fromIterable(userIds).flatMap(this::getUserAsync,5)// 并发度为5.collectList().block(Duration.ofSeconds(30));}// ==================== 回调转同步 ====================publicUsergetUserWithCallback(String userId){CompletableFuture<User> future =newCompletableFuture<>();getUserAsync(userId).subscribe( future::complete, future::completeExceptionally);try{return future.get(10,TimeUnit.SECONDS);}catch(InterruptedException e){Thread.currentThread().interrupt();thrownewRuntimeException("Interrupted", e);}catch(TimeoutException e){thrownewTimeoutException("Timeout waiting for response");}catch(ExecutionException e){thrownewRuntimeException("Execution failed", e);}}// ==================== 响应式到阻塞的转换器 ====================@ComponentpublicstaticclassReactiveToBlockingConverter{privatefinalScheduler scheduler =Schedulers.boundedElastic();public<T>Tconvert(Mono<T> mono,Duration timeout){return mono .subscribeOn(scheduler)// 在指定调度器上执行.block(timeout);}public<T>List<T>convert(Flux<T> flux,Duration timeout){return flux .subscribeOn(scheduler).collectList().block(timeout);}}}

4.2 Spring MVC控制器中的WebClient使用

@RestController@RequestMapping("/api/users")@Slf4jpublicclassUserController{privatefinalWebClient webClient;privatefinalReactiveToBlockingConverter converter;publicUserController(WebClient.Builder webClientBuilder,ReactiveToBlockingConverter converter){this.webClient = webClientBuilder.build();this.converter = converter;}// ==================== 传统Spring MVC控制器 ====================@GetMapping("/{userId}")publicResponseEntity<User>getUser(@PathVariableString userId){// 在MVC控制器中使用WebClient（阻塞方式）try{User user = webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).block(Duration.ofSeconds(10));returnResponseEntity.ok(user);}catch(RuntimeException e){ log.error("Failed to get user", e);returnResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).build();}}// ==================== 异步控制器 ====================@GetMapping("/async/{userId}")publicCompletableFuture<ResponseEntity<User>>getUserAsync(@PathVariableString userId){return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).map(ResponseEntity::ok).onErrorReturn(ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).build()).toFuture();}// ==================== 流式响应 ====================@GetMapping(value ="/stream", produces =MediaType.TEXT_EVENT_STREAM_VALUE)publicFlux<User>streamUsers(@RequestParamList<String> userIds){returnFlux.fromIterable(userIds).delayElements(Duration.ofMillis(100)).flatMap(userId -> webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class));}// ==================== 混合模式：内部响应式，外部阻塞 ====================@GetMapping("/hybrid/{userId}")publicUsergetUserHybrid(@PathVariableString userId){// 使用转换器进行响应式到阻塞的转换Mono<User> userMono = webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).doOnNext(user -> log.info("Retrieved user: {}", user)).doOnError(e -> log.error("Failed to retrieve user", e));return converter.convert(userMono,Duration.ofSeconds(10));}}

五、测试策略迁移

5.1 单元测试对比

@SpringBootTest@AutoConfigureWebTestClient@Slf4jclassWebClientMigrationTest{@MockBeanprivateRestTemplate restTemplate;// 如果还有遗留代码@AutowiredprivateWebTestClient webTestClient;privateWebClient webClient;privateMockWebServer mockWebServer;@BeforeEachvoidsetUp()throwsIOException{ mockWebServer =newMockWebServer(); mockWebServer.start(); webClient =WebClient.builder().baseUrl(mockWebServer.url("/").toString()).build();}@AfterEachvoidtearDown()throwsIOException{ mockWebServer.shutdown();}// ==================== RestTemplate测试（旧方式） ====================@Deprecated@TestvoidtestRestTemplate(){// 模拟RestTemplate响应User mockUser =newUser("123","John Doe");when(restTemplate.getForObject(anyString(),eq(User.class),anyString())).thenReturn(mockUser);// 测试使用RestTemplate的服务// ...}// ==================== WebClient测试（新方式） ====================@TestvoidtestWebClient_Simple(){// 设置MockWebServer响应User mockUser =newUser("123","John Doe"); mockWebServer.enqueue(newMockResponse().setResponseCode(200).setBody("{\"id\":\"123\",\"name\":\"John Doe\"}").addHeader("Content-Type","application/json"));// 执行WebClient调用Mono<User> result = webClient.get().uri("/users/123").retrieve().bodyToMono(User.class);// 验证StepVerifier.create(result).expectNext(mockUser).verifyComplete();// 验证请求RecordedRequest request = mockWebServer.takeRequest();assertEquals("/users/123", request.getPath());assertEquals("GET", request.getMethod());}@TestvoidtestWebClient_Error(){// 模拟错误响应 mockWebServer.enqueue(newMockResponse().setResponseCode(404).setBody("Not Found"));Mono<User> result = webClient.get().uri("/users/999").retrieve().onStatus(status -> status ==HttpStatus.NOT_FOUND, response ->Mono.error(newUserNotFoundException("User not found"))).bodyToMono(User.class);StepVerifier.create(result).expectError(UserNotFoundException.class).verify();}@TestvoidtestWebClient_Timeout(){// 模拟延迟响应 mockWebServer.enqueue(newMockResponse().setResponseCode(200).setBody("{\"id\":\"123\",\"name\":\"John Doe\"}").setBodyDelay(2,TimeUnit.SECONDS));// 2秒延迟WebClient timeoutClient =WebClient.builder().baseUrl(mockWebServer.url("/").toString()).clientConnector(newReactorClientHttpConnector(HttpClient.create().responseTimeout(Duration.ofSeconds(1)))).build();Mono<User> result = timeoutClient.get().uri("/users/123").retrieve().bodyToMono(User.class);StepVerifier.create(result).expectError(ReadTimeoutException.class).verify();}@TestvoidtestWebClient_Concurrent(){// 模拟多个响应for(int i =1; i <=10; i++){ mockWebServer.enqueue(newMockResponse().setResponseCode(200).setBody(String.format("{\"id\":\"%d\",\"name\":\"User %d\"}", i, i)));}Flux<User> result =Flux.range(1,10).flatMap(i -> webClient.get().uri("/users/{id}", i).retrieve().bodyToMono(User.class),3);// 并发度为3StepVerifier.create(result).expectNextCount(10).verifyComplete();}@TestvoidtestWebTestClient_Integration(){// 使用WebTestClient测试控制器 webTestClient.get().uri("/api/users/123").exchange().expectStatus().isOk().expectBody(User.class).value(user ->{assertEquals("123", user.getId());assertEquals("John Doe", user.getName());});}}

5.2 集成测试工具类

@Component@Slf4jpublicclassWebClientTestUtils{/** * 创建测试用的WebClient，指向MockWebServer */publicstaticWebClientcreateTestWebClient(MockWebServer mockWebServer){returnWebClient.builder().baseUrl(mockWebServer.url("/").toString()).clientConnector(newReactorClientHttpConnector(HttpClient.create().wiretap(true)))// 启用网络日志.filter(logRequest()).filter(logResponse()).build();}privatestaticExchangeFilterFunctionlogRequest(){returnExchangeFilterFunction.ofRequestProcessor(clientRequest ->{ log.info("Request: {} {}", clientRequest.method(), clientRequest.url()); clientRequest.headers().forEach((name, values)-> values.forEach(value -> log.debug("{}: {}", name, value)));returnMono.just(clientRequest);});}privatestaticExchangeFilterFunctionlogResponse(){returnExchangeFilterFunction.ofResponseProcessor(clientResponse ->{ log.info("Response: {}", clientResponse.statusCode());returnMono.just(clientResponse);});}/** * 模拟JSON响应 */publicstaticvoidmockJsonResponse(MockWebServer server,Object response)throwsJsonProcessingException{ObjectMapper mapper =newObjectMapper();String json = mapper.writeValueAsString(response); server.enqueue(newMockResponse().setResponseCode(200).setBody(json).addHeader("Content-Type","application/json"));}/** * 模拟错误响应 */publicstaticvoidmockErrorResponse(MockWebServer server,int statusCode,String body){ server.enqueue(newMockResponse().setResponseCode(statusCode).setBody(body));}/** * 验证请求 */publicstaticvoidverifyRequest(MockWebServer server,String expectedPath,String expectedMethod)throwsInterruptedException{RecordedRequest request = server.takeRequest();assertEquals(expectedPath, request.getPath());assertEquals(expectedMethod, request.getMethod());}/** * 创建测试调度器（用于控制测试时间） */publicstaticSchedulercreateTestScheduler(){returnSchedulers.newSingle("test-scheduler");}}

六、性能优化与监控

6.1 连接池配置

@Configuration@Slf4jpublicclassWebClientPerformanceConfig{@BeanpublicConnectionProviderconnectionProvider(){// 配置连接池returnConnectionProvider.builder("webclient-connection-pool").maxConnections(500)// 最大连接数.maxIdleTime(Duration.ofSeconds(20))// 最大空闲时间.maxLifeTime(Duration.ofMinutes(5))// 最大生存时间.pendingAcquireTimeout(Duration.ofSeconds(60))// 获取连接超时.pendingAcquireMaxCount(-1)// 最大等待队列大小.evictInBackground(Duration.ofSeconds(120))// 后台清理间隔.metrics(true)// 启用指标.build();}@BeanpublicHttpClienthttpClient(ConnectionProvider connectionProvider){returnHttpClient.create(connectionProvider).compress(true)// 启用压缩.keepAlive(true)// 保持连接.option(ChannelOption.CONNECT_TIMEOUT_MILLIS,5000)// 连接超时.responseTimeout(Duration.ofSeconds(30))// 响应超时.doOnConnected(conn -> conn.addHandlerLast(newReadTimeoutHandler(30))// 读超时.addHandlerLast(newWriteTimeoutHandler(30)))// 写超时.metrics(true,()->MicrometerHttpClientMetricsRecorder.builder().uriMapper(request -> request.resourceUrl()).build());}@BeanpublicWebClientwebClient(HttpClient httpClient){returnWebClient.builder().clientConnector(newReactorClientHttpConnector(httpClient)).codecs(configurer ->{// 配置编解码器 configurer.defaultCodecs().maxInMemorySize(10*1024*1024);// 10MB}).filter(metricsFilter()).build();}privateExchangeFilterFunctionmetricsFilter(){return(request, next)->{long startTime =System.currentTimeMillis();String requestId = UUID.randomUUID().toString();return next.exchange(request).doOnSuccess(response ->{long duration =System.currentTimeMillis()- startTime;Metrics.counter("http.client.requests","method", request.method().name(),"status",String.valueOf(response.statusCode().value()),"outcome","SUCCESS").increment();Metrics.timer("http.client.duration","method", request.method().name()).record(duration,TimeUnit.MILLISECONDS);}).doOnError(error ->{long duration =System.currentTimeMillis()- startTime;Metrics.counter("http.client.requests","method", request.method().name(),"outcome","ERROR").increment(); log.error("Request {} failed after {}ms", requestId, duration, error);});};}// 监控端点@RestController@RequestMapping("/actuator/webclient")@ConditionalOnProperty(name ="management.endpoint.webclient.enabled", havingValue ="true")publicstaticclassWebClientMetricsEndpoint{@AutowiredprivateConnectionProvider connectionProvider;@GetMapping("/connections")publicMap<String,Object>getConnectionMetrics(){Map<String,Object> metrics =newHashMap<>();// 获取连接池指标if(connectionProvider instanceofConnectionProvider.Metrics){ConnectionProvider.Metrics providerMetrics =(ConnectionProvider.Metrics) connectionProvider; metrics.put("totalConnections", providerMetrics.totalConnections()); metrics.put("activeConnections", providerMetrics.activeConnections()); metrics.put("idleConnections", providerMetrics.idleConnections()); metrics.put("pendingConnections", providerMetrics.pendingConnections());}return metrics;}@PostMapping("/connections/evict")publicResponseEntity<?>evictIdleConnections(){if(connectionProvider instanceofConnectionProvider.Metrics){ connectionProvider.disposeLater().block(Duration.ofSeconds(5));returnResponseEntity.ok("Connections evicted");}returnResponseEntity.badRequest().body("Not supported");}}}

6.2 监控配置

# application.yml - 监控配置management:endpoints:web:exposure:include: health,info,metrics,prometheus,webclient metrics:export:prometheus:enabled:truedistribution:percentiles-histogram:http.client.duration:truetags:application: ${spring.application.name}environment: ${spring.profiles.active}# 自定义指标micrometer:metrics:export:atlas:enabled:falsedatadog:enabled:falseinflux:enabled:falseprometheus:enabled:truestackdriver:enabled:false# WebClient特定监控webclient:metrics:enabled:trueuri-mapper: com.example.WebClientUriMapper

七、渐进式迁移策略

7.1 并行运行与逐步替换

@Component@Slf4jpublicclassGradualMigrationStrategy{@Autowired(required =false)privateRestTemplate legacyRestTemplate;@AutowiredprivateWebClient modernWebClient;// ==================== 迁移阶段1：并行运行 ====================@Component@Profile("migration-phase1")publicstaticclassPhase1Migration{// 1. 创建兼容层publicinterfaceUserServiceClient{UsergetUser(String userId);List<User>getUsers();}// 2. RestTemplate实现（旧）@Component@Primary@ConditionalOnProperty(name ="migration.phase", havingValue ="1")publicstaticclassRestTemplateUserClientimplementsUserServiceClient{privatefinalRestTemplate restTemplate;@OverridepublicUsergetUser(String userId){// RestTemplate实现return restTemplate.getForObject("https://api.example.com/users/{id}",User.class, userId);}@OverridepublicList<User>getUsers(){// RestTemplate实现// ...}}// 3. WebClient实现（新）@Component@ConditionalOnProperty(name ="migration.phase", havingValue ="1")publicstaticclassWebClientUserClientimplementsUserServiceClient{privatefinalWebClient webClient;@OverridepublicUsergetUser(String userId){// WebClient实现（阻塞调用）return webClient.get().uri("https://api.example.com/users/{id}", userId).retrieve().bodyToMono(User.class).block(Duration.ofSeconds(10));}@OverridepublicList<User>getUsers(){// WebClient实现// ...}}// 4. 使用代理模式切换实现@Component@Primary@ConditionalOnProperty(name ="migration.phase", havingValue ="1")publicstaticclassMigrationProxyUserClientimplementsUserServiceClient{privatefinalRestTemplateUserClient legacyClient;privatefinalWebClientUserClient modernClient;privatefinalAtomicBoolean useModern =newAtomicBoolean(false);@OverridepublicUsergetUser(String userId){if(useModern.get()){try{return modernClient.getUser(userId);}catch(Exception e){ log.warn("Modern client failed, falling back to legacy", e);return legacyClient.getUser(userId);}}else{User user = legacyClient.getUser(userId);// 异步验证新客户端verifyModernClientAsync(userId);return user;}}privatevoidverifyModernClientAsync(String userId){CompletableFuture.runAsync(()->{try{ modernClient.getUser(userId); log.info("Modern client verified successfully");}catch(Exception e){ log.warn("Modern client verification failed", e);}});}publicvoidswitchToModern(){ useModern.set(true); log.info("Switched to modern WebClient implementation");}}}// ==================== 迁移阶段2：逐步替换 ====================@Component@Profile("migration-phase2")publicstaticclassPhase2Migration{// 使用包装器模式@ComponentpublicclassRestTemplateWrapper{privatefinalWebClient webClient;privatefinalMap<String,RestTemplate> restTemplateMap =newHashMap<>();// 模拟RestTemplate的方法public<T>ResponseEntity<T>exchange(String url,HttpMethod method,HttpEntity<?> requestEntity,Class<T> responseType,Object... uriVariables){// 将RestTemplate调用转换为WebClient调用WebClient.RequestBodySpec requestSpec = webClient.method(method).uri(url, uriVariables);// 复制请求头if(requestEntity.getHeaders()!=null){ requestEntity.getHeaders().forEach((key, values)-> requestSpec.header(key, values.toArray(newString[0])));}// 处理请求体if(requestEntity.getBody()!=null){ requestSpec.bodyValue(requestEntity.getBody());}// 执行请求Mono<ResponseEntity<T>> responseMono = requestSpec.retrieve().toEntity(responseType);// 阻塞获取结果（兼容模式）return responseMono.block(Duration.ofSeconds(30));}// 其他RestTemplate方法的包装...}}// ==================== 迁移阶段3：完全替换 ====================@Component@Profile("migration-phase3")publicstaticclassPhase3Migration{// 1. 删除所有RestTemplate依赖// 2. 将所有服务转换为响应式// 3. 更新测试代码// 4. 性能优化和监控@BeanpublicWebClient.BuilderwebClientBuilder(){returnWebClient.builder().codecs(configurer ->{ configurer.defaultCodecs().maxInMemorySize(10*1024*1024);}).filter(ExchangeFilterFunction.ofRequestProcessor(request ->{// 统一请求处理returnMono.just(ClientRequest.from(request).header("X-Request-ID", UUID.randomUUID().toString()).build());}));}}}

7.2 迁移检查清单

# migration-checklist.ymlmigration-checklist:phase-1-preparation:-"添加spring-boot-starter-webflux依赖"-"创建WebClient配置"-"实现兼容层接口"-"编写并行测试用例"-"配置监控和日志"phase-2-implementation:-"逐个服务迁移"-"更新错误处理逻辑"-"调整超时和重试配置"-"性能基准测试"-"验证功能正确性"phase-3-testing:-"单元测试迁移"-"集成测试更新"-"负载测试验证"-"错误场景测试"-"性能对比测试"phase-4-production:-"监控指标验证"-"逐步流量切换"-"回滚方案准备"-"文档更新"-"团队培训"common-challenges:-"同步到异步的思维转变"-"错误处理机制差异"-"线程池和资源管理"-"测试策略调整"-"调试复杂性增加"best-practices:-"使用StepVerifier进行响应式测试"-"合理配置连接池"-"实现断路器模式"-"添加详细的监控指标"-"保持代码可读性"

八、常见问题解决方案

8.1 问题诊断与解决

@Component@Slf4jpublicclassWebClientTroubleshooting{// ==================== 问题1：内存泄漏 ====================publicvoiddiagnoseMemoryLeak(){// 症状：内存持续增长// 原因：未正确释放资源// 解决方案：// 1. 使用适当的操作符 webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class).doOnNext(data ->process(data)).doOnError(e -> log.error("Error", e)).subscribe();// ❌ 危险的subscribe// 正确的做法：Disposable disposable = webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class).subscribe( data ->process(data), error -> log.error("Error", error));// 在适当的时候取消订阅 disposable.dispose();// 更好的做法：使用响应式控制器@GetMapping("/data")publicMono<Data>getData(){return webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class);}}// ==================== 问题2：线程阻塞 ====================publicvoiddiagnoseThreadBlocking(){// 症状：线程池耗尽// 原因：在响应式链中执行阻塞操作// ❌ 错误示例 webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class).map(data ->{// 阻塞操作！returnblockingDatabaseCall(data);});// ✅ 正确解决方案 webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class).flatMap(data ->Mono.fromCallable(()->blockingDatabaseCall(data)).subscribeOn(Schedulers.boundedElastic()));}// ==================== 问题3：背压处理 ====================publicvoiddiagnoseBackpressure(){// 症状：生产者速度 > 消费者速度// 解决方案：使用背压操作符Flux<Data> dataStream = webClient.get().uri("/api/stream").retrieve().bodyToFlux(Data.class);// 限制处理速率 dataStream .limitRate(10)// 每次请求10个元素.delayElements(Duration.ofMillis(100))// 延迟处理.subscribe( data ->process(data), error -> log.error("Error", error),()-> log.info("Completed"), subscription -> subscription.request(10)// 初始请求);}// ==================== 问题4：调试困难 ====================publicvoidenableDebugging(){// 1. 启用详细日志System.setProperty("reactor.netty.http.client.level","DEBUG");// 2. 添加调试操作符 webClient.get().uri("/api/data").retrieve().bodyToMono(Data.class).log("webclient.request")// 添加日志.doOnSubscribe(s -> log.debug("Subscribed")).doOnNext(d -> log.debug("Received: {}", d)).doOnError(e -> log.error("Error: ", e)).doOnTerminate(()-> log.debug("Terminated"));// 3. 使用Hooks.onOperatorDebug()Hooks.onOperatorDebug();// 4. 启用堆栈跟踪收集Schedulers.enableMetrics();}// ==================== 问题5：配置复杂性 ====================@ConfigurationpublicstaticclassWebClientSimplification{// 创建预配置的WebClient [email protected](){returnWebClient.builder().codecs(configurer ->{ configurer.defaultCodecs().maxInMemorySize(10*1024*1024);}).defaultHeader(HttpHeaders.USER_AGENT,"MyApp/1.0").filter(ExchangeFilterFunction.ofRequestProcessor(request ->{// 添加跟踪IDString traceId = MDC.get("traceId");if(traceId !=null){returnMono.just(ClientRequest.from(request).header("X-Trace-ID", traceId).build());}returnMono.just(request);}));}// 为特定服务创建专用WebClient@Bean("userServiceWebClient")publicWebClientuserServiceWebClient(WebClient.Builder builder){return builder.clone().baseUrl("https://api.users.example.com").defaultHeader("X-Service","user-service").build();}@Bean("paymentServiceWebClient")publicWebClientpaymentServiceWebClient(WebClient.Builder builder){return builder.clone().baseUrl("https://api.payments.example.com").defaultHeader("X-Service","payment-service").build();}}}

总结

RestTemplate迁移到WebClient的关键要点：

  1. 思维转变：从同步阻塞到异步非阻塞的编程模式
  2. 渐进迁移：采用并行运行、逐步替换的策略
  3. 错误处理：利用WebClient强大的错误处理机制
  4. 性能优化：合理配置连接池、超时和重试策略
  5. 监控完善：添加详细的指标和日志
  6. 测试保障：使用StepVerifier等工具进行响应式测试

通过遵循本文的迁移策略和解决方案，可以顺利完成从RestTemplate到WebClient的迁移，并享受到响应式编程带来的性能优势和更好的资源利用率。

Read more

【前端部署在云服务器如何与本地联调--Frp内网穿透】

【前端部署在云服务器如何与本地联调--Frp内网穿透】

苍穹外卖前端部署在云服务器如何与本地联调--Frp内网穿透 * 1. 前言 * 2. FRP是什么 * 3. 解决步骤 * 3.1 在云服务器安装服务端frps，然后开启开机自启（**参考第4部分**） * 3.2 在本地电脑安装客户端fprc（**参考第4部分**） * 4. Frp(C/S)0.64.0各个系统的安装方法 * 4.1 frps安装（Linux）服务端 * 4.2 frpc安装(windows)客户端 * 4.3 frpc安装(==mac==)客户端 * 4.4 frpc安装(Linux)客户端 1. 前言 写这片文章的目的是为了解决上篇苍穹外卖项目的前端部署到云服务器的遗留问题：前端的云服务器的IP是公网IP，而我本地调试的Java后端是内网，前端响应的地址找不到本地的服务器。那么如何让云服务器上的前端项目能够找到后端的对应的地址呢？
前端实战：基于Vue3与免费满血版DeepSeek实现无限滚动+懒加载+瀑布流模块及优化策略

前端实战：基于Vue3与免费满血版DeepSeek实现无限滚动+懒加载+瀑布流模块及优化策略

目录 前端实战：基于Vue3与免费满血版DeepSeek实现无限滚动+懒加载+瀑布流模块及优化策略 一、前言 二、如何使用腾讯云免费满血版deepseek 1、腾讯云大模型知识引擎体验中心 2、体验deepseek联网助手 3、人机交互获取AI支持 三、基于DeepSeek实现无限滚动+懒加载+瀑布流模块 1、无限滚动+懒加载+瀑布流模块的底层逻辑 2、人机交互策略与Deepseek的实现过程 ①虚拟列表管理 ②布局容器初始化 ③动态渲染与销毁机制 ④无线滚动实现 ⑤内存优化策略 四、最终代码呈现 1、组件代码 2、组件用法 五、结语         作者：watermelo37         ZEEKLOG万粉博主、华为云云享专家、阿里云专家博主、腾讯云、支付宝合作作者，全平台博客昵称watermelo37。         一个假装是giser的coder，做不只专注于业务逻辑的前端工程师，Java、Docker、

在自动化脚本中如何在自定义ui中使用webview来无限扩展ui？

在自动化脚本开发中，原生 UI 控件虽能满足基础的界面展示与交互需求，但面对复杂的页面逻辑、动态的内容渲染以及个性化的交互设计时，其扩展性会受到一定限制。WebView 控件能够将网页的灵活开发特性与自动化脚本的原生能力深度融合，实现 UI 的无限扩展。本文将从 WebView 的集成原理、与自动化脚本的无缝交互方式出发，结合完整的 Demo 源码，详细讲解如何在UI 中高效集成 WebView，让 H5 页面与原生自动化脚本协同工作，打造更灵活、更强大的自动化交互界面。 一、WebView 核心能力与集成前提 1.1 WebView 的核心价值  WebView 控件并非简单的网页加载容器，而是打通了原生自动化脚本与H5 网页的双向通信通道，其核心价值体现在三个方面： 1. UI 扩展无限化：借助 H5 的生态优势，实现原生 UI 难以开发的复杂界面，如数据可视化图表、动态表单、
实验三 Windows Server 2022/2025 搭建 Web 服务器实验指导书

实验三 Windows Server 2022/2025 搭建 Web 服务器实验指导书

作者：非凡大爹｜版本：v1｜日期：2026-03-30｜DocID：CN-LAB-2026-03-WEB-1-LG-V1 原创声明：本文为非凡大爹原创，首发于ZEEKLOG，转载或引用请注明出处。 一、实验基本信息 课程名称： Windows 网络管理 / 网络操作系统 / 服务器配置与管理 实验名称： Windows Server 2022/2025 搭建 Web 服务器 实验性质： 验证性 + 应用性实验 实验类别： 综合配置实验 建议学时： 2 学时 实验方式： 学生独立操作 + 结果验证 二、实验目的 1. 知识目标 理解 Web 服务器的基本作用，了解网站从“本地网页文件”到“网络可访问服务”的基本发布过程，