如果想了解關閉socket的細節，可以通過HttpClientConnection.close()繼續往下跟蹤，最終會看到真正關閉socket的代碼在BHttpConnectionBase中。

/**
* This class serves as a base for all {@link HttpConnection} implementations and provides
* functionality common to both client and server HTTP connections.
*
* @since 4.0
*/
public class BHttpConnectionBase implements HttpConnection, HttpInetConnection {
... ...
@Override
public void close() throws IOException {
final Socket socket = this.socketHolder.getAndSet(null);
if (socket != null) {
try {
this.inbuffer.clear();
this.outbuffer.flush();
try {
try {
socket.shutdownOutput();
} catch (final IOException ignore) {
}
try {
socket.shutdownInput();
} catch (final IOException ignore) {
}
} catch (final UnsupportedOperationException ignore) {
// if one isnt supported, the other one isnt either
}
} finally {
socket.close();
}
}
}
... ...
}


為什麼說調用了EntityUtils的部分方法後，就不需要再顯示地關閉流呢？看下它的源碼就明白了。

/**
* Static helpers for dealing with {@link HttpEntity}s.
*
* @since 4.0
*/
public final class EntityUtils {
/**
* Ensures that the entity content is fully consumed and the content stream, if exists,
* is closed.
*
* @param entity the entity to consume.
* @throws IOException if an error occurs reading the input stream
*
* @since 4.1
*/
public static void consume(final HttpEntity entity) throws IOException {
if (entity == null) {
return;
}
if (entity.isStreaming()) {
final InputStream instream = entity.getContent();
if (instream != null) {
instream.close();
}
}
}

... ...
}

(3)HttpClient進階用法

在高並發場景下，使用連接池有效復用已經建立的連接是非常必要的。如果每次http請求都重新建立連接，那麼底層的socket連接每次通過3次握手創建和4次握手斷開連接將是一筆非常大的時間開銷。

private static volatile CloseableHttpClient instance;

static {
PoolingHttpClientConnectionManager cm = new PoolingHttpClientConnectionManager();
// Increase max total connection to 200
cm.setMaxTotal(200);
// Increase default max connection per route to 20
cm.setDefaultMaxPerRoute(20);
RequestConfig requestConfig = RequestConfig.custom()
.setConnectTimeout(1000)
.setSocketTimeout(1000)
.setConnectionRequestTimeout(1000)
.build();
instance = HttpClients.custom()
.setConnectionManager(cm)
.setDefaultRequestConfig(requestConfig)
.build();

}

官方同時建議我們在後臺起一個定時清理無效連接的線程，因為某些連接建立後可能由於服務端單方面斷開連接導致一個不可用的連接一直佔用著資源，而HttpClient框架又不能百分之百保證檢測到這種異常連接並做清理，因此需要自給自足，按照如下方式寫一個空閑連接清理線程在後臺運行。

public class IdleConnectionMonitorThread extends Thread {
private final HttpClientConnectionManager connMgr;
private volatile boolean shutdown;
Logger logger = LoggerFactory.getLogger(IdleConnectionMonitorThread.class);

public IdleConnectionMonitorThread(HttpClientConnectionManager connMgr) {
super();
this.connMgr = connMgr;
}

@Override
public void run() {
try {
while (!shutdown) {
synchronized (this) {
wait(5000);
// Close expired connections
connMgr.closeExpiredConnections();
// Optionally, close connections
// that have been idle longer than 30 sec
connMgr.closeIdleConnections(30, TimeUnit.SECONDS);
} }
} catch (InterruptedException ex) {
logger.error("unknown exception", ex);
// terminate
}
}

public void shutdown() {
shutdown = true;
synchronized (this) {
notifyAll();
}
}
}

我們討論到的幾個核心類的依賴關係如下：

HttpClient作為大家常用的工具，看似簡單，但是其中卻有很多隱藏的細節值得探索。

本文作者：閑魚技術-峯明

原文鏈接

更多技術乾貨敬請關注云棲社區知乎機構號：阿里云云棲社區 - 知乎

本文為雲棲社區原創內容，未經允許不得轉載。

推薦閱讀：