小强哥博客

小强哥,小强哥博客,技术大咖

ranong项目总结-HttpClient-RetryHandler

目前的项目接口都是http,因此在java项目中使用apache httpclient进行数据传输、访问。

目前程序中涉及到需要callback操作,product需要被动的接收consume的处理状态,为了最大程度的能够callback成功因此consume在http调用出现问题(如:服务不可用、异常、超时)情况下需要进行重试(retry request),在这里我列举出我找到的retry方案,有些成功有些不成功。

我是用的httpclient版本是4.5.2。关于retry功能我在网上也找了不少的资料,但是都不是我对应的httpclient版本,大多是过时的。

在httpclient版本4.5.2提供了以下几种retry方案:

StandardHttpRequestRetryHandler

这种方案没有测试通过,StandardHttpRequestRetryHandler实际上是DefaultHttpRequestRetryHandler的子类,这是官方提供的一个标准的retry方案,为了保证幂等性约定resetful接口必须是GET, HEAD, PUT, DELETE, OPTIONS, and TRACE中的一种,如下,是我定义的httpclient pool,

public static CloseableHttpClient getHttpClient() {
		PoolingHttpClientConnectionManager cm = new PoolingHttpClientConnectionManager();
		cm.setMaxTotal(MAX_TOTAL);
		cm.setDefaultMaxPerRoute(MAX_PERROUTE);
		CloseableHttpClient httpClient = HttpClients
				.custom()
				.setRetryHandler(new StandardHttpRequestRetryHandler())
				.setConnectionManager(cm)
				.build();
		return httpClient;
	}

如下是我的测试代码,

@Test
	public void test6(){
		HttpPost httpPost=new HttpPost("http://127.0.0.1:8080/testjobs1");
		try {
			rsp=httpClient.execute(httpPost);
			log.info(">> {}",rsp.getStatusLine().getStatusCode());
		} catch (Exception e) {
			log.error(e.getMessage(),e);
		}finally{
			HttpUtil.close(rsp);
		}
	}

运行测试,当url错误、后台报错、后台超时等情况的时候不能进行retry,因此放弃了此方案。

DefaultHttpRequestRetryHandler

这种方案没有测试通过,和上面的StandardHttpRequestRetryHandler类似,它提供了一种默认的retry方案,并没有像StandardHttpRequestRetryHandler一样约定接口必须是冥等的,如下,是我定义的httpclient pool,

public static CloseableHttpClient getHttpClient() {
		PoolingHttpClientConnectionManager cm = new PoolingHttpClientConnectionManager();
		cm.setMaxTotal(MAX_TOTAL);
		cm.setDefaultMaxPerRoute(MAX_PERROUTE);
		CloseableHttpClient httpClient = HttpClients
				.custom()
				.setRetryHandler(new DefaultHttpRequestRetryHandler())
				.setConnectionManager(cm)
				.build();
		return httpClient;
	}

如下是我的测试代码,

@Test
	public void test6(){
		HttpPost httpPost=new HttpPost("http://127.0.0.1:8080/testjobs1");
		try {
			rsp=httpClient.execute(httpPost);
			log.info(">> {}",rsp.getStatusLine().getStatusCode());
		} catch (Exception e) {
			log.error(e.getMessage(),e);
		}finally{
			HttpUtil.close(rsp);
		}
	}

依然没有达到希望的效果。

HttpRequestRetryHandler

可以实现,但是不够完美。在官方文档有这么一段,如下,

HttpRequestRetryHandler myRetryHandler = new HttpRequestRetryHandler() {

    public boolean retryRequest(
            IOException exception,
            int executionCount,
            HttpContext context) {
        if (executionCount >= 5) {
            // Do not retry if over max retry count
            return false;
        }
        if (exception instanceof InterruptedIOException) {
            // Timeout
            return false;
        }
        if (exception instanceof UnknownHostException) {
            // Unknown host
            return false;
        }
        if (exception instanceof ConnectTimeoutException) {
            // Connection refused
            return false;
        }
        if (exception instanceof SSLException) {
            // SSL handshake exception
            return false;
        }
        HttpClientContext clientContext = HttpClientContext.adapt(context);
        HttpRequest request = clientContext.getRequest();
        boolean idempotent = !(request instanceof HttpEntityEnclosingRequest);
        if (idempotent) {
            // Retry if the request is considered idempotent
            return true;
        }
        return false;
    }

};
CloseableHttpClient httpclient = HttpClients.custom()
        .setRetryHandler(myRetryHandler)
        .build();

自定义retry实现,这比较灵活,可以根据异常自定义retry机制以及重试次数,并且可以拿到返回信息,如下,是我定义的httpclient pool,

public static CloseableHttpClient getHttpClient() {
		HttpRequestRetryHandler retryHandler = new HttpRequestRetryHandler() {
		    public boolean retryRequest(
		            IOException exception,
		            int executionCount,
		            HttpContext context) {
		        if (executionCount >= 5) {
		            // Do not retry if over max retry count
		            return false;
		        }
		        if (exception instanceof InterruptedIOException) {
		            // Timeout
		            return false;
		        }
		        if (exception instanceof UnknownHostException) {
		            // Unknown host
		            return false;
		        }
		        if (exception instanceof ConnectTimeoutException) {
		            // Connection refused
		            return false;
		        }
		        if (exception instanceof SSLException) {
		            // SSL handshake exception
		            return false;
		        }
		        HttpClientContext clientContext = HttpClientContext.adapt(context);
		        HttpRequest request = clientContext.getRequest();
		        boolean idempotent = !(request instanceof HttpEntityEnclosingRequest);
		        if (idempotent) {
		            // Retry if the request is considered idempotent
		            return true;
		        }
		        return false;
		    }
		};
		PoolingHttpClientConnectionManager cm = new PoolingHttpClientConnectionManager();
		cm.setMaxTotal(MAX_TOTAL);
		cm.setDefaultMaxPerRoute(MAX_PERROUTE);
		CloseableHttpClient httpClient = HttpClients
				.custom()
				.setRetryHandler(retryHandler)
				.setConnectionManager(cm)
				.build();
		return httpClient;
	}

如下是我的测试代码,

@Test
	public void test6(){
		HttpPost httpPost=new HttpPost("http://127.0.0.1:8080/testjobs1");
		try {
			rsp=httpClient.execute(httpPost);
			log.info(">> {}",rsp.getStatusLine().getStatusCode());
		} catch (Exception e) {
			log.error(e.getMessage(),e);
		}finally{
			HttpUtil.close(rsp);
		}
	}

这种方案,可以实现retry,并且可以根据我的需求进行retry,如:retry count,但是就不能控制retry时间的间隔,也只好放弃了,继续寻找找到了下面这个ServiceUnavailableRetryStrategy。

ServiceUnavailableRetryStrategy

可以实现,满足需求,这具有HttpRequestRetryHandler的所有有点,并且可以自定义retry时间的间隔,如下,是我定义的httpclient pool,

public static CloseableHttpClient getHttpClient() {
		ServiceUnavailableRetryStrategy serviceUnavailableRetryStrategy = new ServiceUnavailableRetryStrategy() {
			/**
			 * retry逻辑
			 */
			@Override
			public boolean retryRequest(HttpResponse response, int executionCount, HttpContext context) {
				if (executionCount <= 3)
					return true;
				else
					return false;
			}

			/**
			 * retry间隔时间
			 */
			@Override
			public long getRetryInterval() {
				return 2000;
			}
		};
		PoolingHttpClientConnectionManager cm = new PoolingHttpClientConnectionManager();
		cm.setMaxTotal(MAX_TOTAL);
		cm.setDefaultMaxPerRoute(MAX_PERROUTE);
		CloseableHttpClient httpClient = HttpClients.custom().setRetryHandler(new DefaultHttpRequestRetryHandler())
				.setConnectionManager(cm).build();
		return httpClient;
	}

如下是我的测试代码,

@Test
	public void test6(){
		HttpPost httpPost=new HttpPost("http://127.0.0.1:8080/testjobs1");
		try {
			rsp=httpClient.execute(httpPost);
			log.info(">> {}",rsp.getStatusLine().getStatusCode());
		} catch (Exception e) {
			log.error(e.getMessage(),e);
		}finally{
			HttpUtil.close(rsp);
		}
	}