Thrift 提供了如圖五種模式：TSimpleServer、TNonblockingServer、THsHaServer、TThreadPoolServer、TThreadSelectorServer

TSimpleServer、TThreadPoolServer 屬于阻塞模型

TNonblockingServer、THsHaServer、TThreadedSelectorServer 屬于非阻塞模型

TServer

TServer 為抽象類

public static class Args extends AbstractServerArgs<Args> {public Args(TServerTransport transport) {super(transport);}
}public static abstract class AbstractServerArgs<T extends AbstractServerArgs<T>> {final TServerTransport serverTransport;// 處理層工廠
  TProcessorFactory processorFactory;// 傳輸層工廠TTransportFactory inputTransportFactory = new TTransportFactory();TTransportFactory outputTransportFactory = new TTransportFactory();// 協議層工廠TProtocolFactory inputProtocolFactory = new TBinaryProtocol.Factory();TProtocolFactory outputProtocolFactory = new TBinaryProtocol.Factory();
}

TServer 定義的對外方法

/*** The run method fires up the server and gets things going.*/public abstract void serve();
/*** Stop the server. This is optional on a per-implementation basis. Not* all servers are required to be cleanly stoppable.*/public void stop() {}

stop 并不是每個服務都需要優雅的退出，所以沒有定義為抽象方法

抽象方法 serve() 由具體的 TServer 實例實現

TSimpleServer

TSimpleServer 實現比較簡單，是單線程阻塞模型，只適合測試開發使用

serve 方法源碼分析

public void serve() {// 啟動監聽 socket
  serverTransport.listen();// 設置服務狀態setServing(true);// 不斷的等待與處理 socket 請求while(!stopped) {// accept 一個業務 socket 請求client = serverTransport_.accept();if (client != null) {// 通過工廠獲取 server 定義的處理層、傳輸層和協議層processor = processorFactory_.getProcessor(client);inputTransport = inputTransportFactory_.getTransport(client);outputTransport = outputTransportFactory_.getTransport(client);inputProtocol = inputProtocolFactory_.getProtocol(inputTransport);outputProtocol = outputProtocolFactory_.getProtocol(outputTransport);if (eventHandler_ != null) {connectionContext = eventHandler_.createContext(inputProtocol, outputProtocol);}// 阻塞式處理while (true) {// 處理請求事件if (eventHandler_ != null) {eventHandler_.processContext(connectionContext, inputTransport, outputTransport);}// 如果處理層為異步，則退出if(!processor.process(inputProtocol, outputProtocol)) {break;}}}// 關閉
    eventHandler_.deleteContext(connectionContext, inputProtocol, outputProtocol);inputTransport.close();outputTransport.close();setServing(false);}
}

TSimpleServer 工作圖

draw.io

TThreadPoolServer

ThreadPoolServer 解決了 TSimple 不支持并發和多連接的問題，引入了線程池

與 TSimple 相同，主線程負責阻塞式監聽 socket，而剩下的業務處理則全部交由線程池去處理

public void serve() {// 主線程啟動監聽 socket
  serverTransport_.listen();// 設置服務狀態stopped_ = false;setServing(true);// 等待并處理 socket 請求while (!stopped_) {TTransport client = serverTransport_.accept();// Runnable run 邏輯與 TSimpleServer 類似WorkerProcess wp = new WorkerProcess(client);int retryCount = 0;long remainTimeInMillis = requestTimeoutUnit.toMillis(requestTimeout);while(true) {// 交由線程池來處理
        executorService_.execute(wp);break;} }executorService_.shutdown();setServing(false);
}

TThreadPoolServer 的缺點：

處理能力的好壞受限于線程池的設置

TNoblockingServer

TNoblockingServer 是單線程工作，但該模式采用了 NIO，所有的 socket 被注冊到 selector 中，通過一個線程循環 selector 來監控所有 socket，當有就緒的 socket 時，根據不同的請求做不同的動作（讀取、寫入數據或 accept 新連接）

TNoblockingServer 的 serve 方法在其父類 AbstractNonblockingServer 中定義

/*** Begin accepting connections and processing invocations.* 開始接受并處理調用*/
public void serve() {// start any IO threads// 注冊一些監聽 socket 的線程到 selector 中if (!startThreads()) {return;}// start listening, or exit// 開始監聽if (!startListening()) {return;}// 設置服務狀態setServing(true);// this will block while we serve// TNonblocking 中實現為 selectAcceptThread_.join(); // 主線程等待 selectAcceptThread 執行完畢// SelectAcceptThread 的 run 方法為 select();// 取出一個就緒的 socket
  waitForShutdown();setServing(false);// do a little cleanup
  stopListening();
}// SelectAcceptThread run 方法
public void run() {while (!stopped_) {select();processInterestChanges();}for (SelectionKey selectionKey : selector.keys()) {cleanupSelectionKey(selectionKey);}
}// SelectAcceptThread Select 過程
private void select() {try {// wait for io events.// NIO 取出一個
    selector.select();// process the io events we receivedIterator<SelectionKey> selectedKeys = selector.selectedKeys().iterator();// 遍歷就緒的 socketwhile (!stopped_ && selectedKeys.hasNext()) {SelectionKey key = selectedKeys.next();selectedKeys.remove();// if the key is marked Accept, then it has to be the server// transport.// accept 新 socket 并將其注冊到 selector 中if (key.isAcceptable()) {handleAccept();} else if (key.isReadable()) {// deal with reads// 處理讀數據的 socket 請求
        handleRead(key);} else if (key.isWritable()) {// deal with writes// 處理寫數據的 socket 請求
        handleWrite(key);} else {LOGGER.warn("Unexpected state in select! " + key.interestOps());}}} catch (IOException e) {LOGGER.warn("Got an IOException while selecting!", e);}
}// 接收新的連接
private void handleAccept() throws IOException {SelectionKey clientKey = null;TNonblockingTransport client = null;// accept the connectionclient = (TNonblockingTransport)serverTransport.accept();// 注冊到 selector 中clientKey = client.registerSelector(selector, SelectionKey.OP_READ);// add this key to the mapFrameBuffer frameBuffer = createFrameBuffer(client, clientKey, SelectAcceptThread.this);clientKey.attach(frameBuffer);
}

TNonblockingServer 模式的缺點：

其還是采用單線程順序來完成，當業務處理比較復雜耗時，該模式的效率將會下降

TNonblockingServer 工作圖：

draw.io

THsHaServer

THsHaServer 是 TNoblockingServer 的子類，處理邏輯基本相同，不同的是，在處理讀取請求時，THsHaServer 將處理過程交由線程池來完成，主線程直接返回進行下一次循環，提高了效率

THsHaServer 模式的缺點：

其主線程需要完成對所有 socket 的監聽一級數據的寫操作，當大請求量時，效率較低

TThreadedSelectorServer

TThreadedSelectorServer 是 Thrift 目前提供的最高級模式，生產環境的首選，其對 TNonblockingServer 進行了擴展

TThreadedSelectorServer 源碼中一些關鍵的屬性

public static class Args extends AbstractNonblockingServerArgs<Args> {// 在已接收的連接中選擇線程的個數public int selectorThreads = 2;// 執行線程池 ExecutorService 的線程個數private int workerThreads = 5;// 執行請求具體任務的線程池private ExecutorService executorService = null;
}
// The thread handling all accepts
private AcceptThread acceptThread;
// Threads handling events on client transports
private final Set<SelectorThread> selectorThreads = new HashSet<SelectorThread>();
// This wraps all the functionality of queueing and thread pool management
// for the passing of Invocations from the selector thread(s) to the workers
// (if any).
private final ExecutorService invoker;
/*** 循環模式的負載均衡器，用于為新的連接選擇 SelectorThread*/
protected static class SelectorThreadLoadBalancer {}

AcceptThread 線程對象，用于監聽 socket 的新連接
多個 SelectorThread 線程對象，用于處理 socket 的讀寫操作
一個負載均衡對象 SelectorThreadLoadBalancer，用于決定將 AcceptThread 接收到的 socket 請求分配給哪個 SelectorThread 線程
SelectorThread 線程執行過讀寫操作后，通過 ExecutorService 線程池來完成此次調用的具體執行

SelectorThread 對象源碼解析

/*** 多個 SelectorThread 負責處理 socket 的 I/O 操作*/
protected class SelectorThread extends AbstractSelectThread {/*** The work loop. Handles selecting (read/write IO), dispatching, and* managing the selection preferences of all existing connections.* 選擇（處理 socket 的網絡讀寫 IO），分配和管理現有連接*/public void run() {while (!stopped_) {select();}}private void select() {// process the io events we receivedIterator<SelectionKey> selectedKeys = selector.selectedKeys().iterator();while (!stopped_ && selectedKeys.hasNext()) {SelectionKey key = selectedKeys.next();selectedKeys.remove();// skip if not validif (!key.isValid()) {cleanupSelectionKey(key);continue;}if (key.isReadable()) {// deal with reads
        handleRead(key);} else if (key.isWritable()) {// deal with writes
        handleWrite(key);} else {LOGGER.warn("Unexpected state in select! " + key.interestOps());}}}
}

AcceptThread 對象源碼解析

/*** 在服務器傳輸中選擇線程（監聽 socket 請求）并向 IO 選擇器（SelectorThread）提供新連接*/
protected class AcceptThread extends Thread {// The listen socket to accept onprivate final TNonblockingServerTransport serverTransport;private final Selector acceptSelector;// 負載均衡器，決定將連接分配給哪個 SelectorThreadprivate final SelectorThreadLoadBalancer threadChooser;public void run() {while (!stopped_) {select();}}private void select() {// process the io events we receivedIterator<SelectionKey> selectedKeys = acceptSelector.selectedKeys().iterator();while (!stopped_ && selectedKeys.hasNext()) {SelectionKey key = selectedKeys.next();selectedKeys.remove();// 處理接收的新情求if (key.isAcceptable()) {handleAccept();} else {LOGGER.warn("Unexpected state in select! " + key.interestOps());}}}/*** Accept a new connection.*/private void handleAccept() {final TNonblockingTransport client = doAccept();if (client != null) {// 從負載均衡器中，獲取 SelectorThread 線程final SelectorThread targetThread = threadChooser.nextThread();if (args.acceptPolicy == Args.AcceptPolicy.FAST_ACCEPT || invoker == null) {doAddAccept(targetThread, client);} else {// FAIR_ACCEPTinvoker.submit(new Runnable() {public void run() {// 將選擇到的線程和連接放入 線程池 處理// 用 targetThread 線程取處理一個給接受的鏈接 client，如果新連接的隊列處于滿的狀態，則將處于阻塞狀態
            doAddAccept(targetThread, client);}});}}}private TNonblockingTransport doAccept() {return (TNonblockingTransport) serverTransport.accept();}// 用 targetThread 線程取處理一個給接受的鏈接 client，如果新連接的隊列處于滿的狀態，則將處于阻塞狀態private void doAddAccept(SelectorThread thread, TNonblockingTransport client) {if (!thread.addAcceptedConnection(client)) {client.close();}}
}