Hadoop datanode启动源码解析

dn不多做介绍，启动流程相对简单，通过main方法开始，一步一步去看源码，本文源码解析只展示重要方法

顺便求志同道合者，致力于大数据存储与计算
有无交流群共同学习（Hadoop源码/性能调优/生产环境真实场景）

DN流程图

源码解析

定位到DataNode.java，DN也是一个main程序

public static void main(String args[]) {
    if (DFSUtil.parseHelpArgument(args, DataNode.USAGE, System.out, true)) {
      System.exit(0);
    }
    secureMain(args, null);
  }

进入secureMain(args, null)，创建DN

public static void secureMain(String args[], SecureResources resources) {
    int errorCode = 0;
    try {
        StringUtils.startupShutdownMessage(DataNode.class, args, LOG);
        //创建DN
        DataNode datanode = createDataNode(args, null, resources);
        if (datanode != null) {
            //阻塞
            datanode.join();
        }
    }
}

进入createDataNode(args, null, resources)，创建DN

/** Instantiate & Start a single datanode daemon and wait for it to finish.
   *  If this thread is specifically interrupted, it will stop waiting.
   */
@VisibleForTesting
@InterfaceAudience.Private
  public static DataNode createDataNode(String args[], Configuration conf,
                                        SecureResources resources) throws IOException {
    //初始化DN
    DataNode dn = instantiateDataNode(args, conf, resources);
    if (dn != null) {
        //启动DN
        dn.runDatanodeDaemon();
    }
    return dn;
}

进入instantiateDataNode(args, conf, resources)，创建DN

/** Instantiate a single datanode object, along with its secure resources. 
   * This must be run by invoking{@link DataNode#runDatanodeDaemon()} 
   * subsequently. 
   */
public static DataNode instantiateDataNode(String args [], Configuration conf,
                                           SecureResources resources) throws IOException {

    //...配置相关

	//实例
    return makeInstance(dataLocations, conf, resources);
}

进入makeInstance(dataLocations, conf, resources)，检查目录、创建DN

static DataNode makeInstance(Collection<StorageLocation> dataDirs,
                             Configuration conf, SecureResources resources) throws IOException {
    //...
    //目录检查，返回可用的磁盘目录列表
    locations = storageLocationChecker.check(conf, dataDirs);
    //构造DN
    return new DataNode(conf, locations, storageLocationChecker, resources);
}

进入new DataNode(conf, locations, storageLocationChecker, resources)，配置、创建DN

DataNode(final Configuration conf,
         final List<StorageLocation> dataDirs,
         final StorageLocationChecker storageLocationChecker,
         final SecureResources resources) throws IOException {
    //...配置

    //启动DN
    startDataNode(dataDirs, resources);

    //...
}

进入startDataNode(dataDirs, resources)，初始化各种线程服务，存储DataStorage/读写相关DataXceiver/RPC相关IpcServer/块池管理blockPoolManager等多个服务

void startDataNode(List<StorageLocation> dataDirectories,
                     SecureResources resources
                     ) throws IOException {

    // 初始化DataStorage
    //DataStorage：管理与组织磁盘存储目录，如current,previous,detach,tmp等；
    //在DataNode数据目录，可以看到一些current tmp,rbw或者finalized文件夹
    //FsDatasetImpl：管理组织数据块和元数据文件
    storage = new DataStorage();
    
    // 注册JMX
    registerMXBean();
    // 初始化DataXceiverServer（流式通信），DataNode#runDatanodeDaemon()中启动
    initDataXceiver();
    // 启动InfoServer
    startInfoServer();
    // 启动JVMPauseMonitor（反向监控JVM情况，可通过JMX查询）
    pauseMonitor = new JvmPauseMonitor();
    pauseMonitor.init(getConf());
    pauseMonitor.start();
  
    initIpcServer();
	//ec
    ecWorker = new ErasureCodingWorker(getConf(), this);
    blockRecoveryWorker = new BlockRecoveryWorker(this);

	//创建块池管理对象
    blockPoolManager = new BlockPoolManager(this);
    //刷新NN
    blockPoolManager.refreshNamenodes(getConf());
  }

进入blockPoolManager.refreshNamenodes(getConf())，获取NN RPC地址，完成NameNodes的刷新

void refreshNamenodes(Configuration conf)
    throws IOException {
    //NN RPC地址
    try {
        newAddressMap =
            DFSUtil.getNNServiceRpcAddressesForCluster(conf);
        newLifelineAddressMap =
            DFSUtil.getNNLifelineRpcAddressesForCluster(conf);
    } 

    synchronized (refreshNamenodesLock) {
        doRefreshNamenodes(newAddressMap, newLifelineAddressMap);
    }
}

进入doRefreshNamenodes(newAddressMap, newLifelineAddressMap)，完成NameNodes的刷新

private void doRefreshNamenodes(
    Map<String, Map<String, InetSocketAddress>> addrMap,
    Map<String, Map<String, InetSocketAddress>> lifelineAddrMap)
    throws IOException {

    //确认NN
    //移除不存在NN
    //遍历联邦，启动新NN，启动BPOfferService，实际启动BPServiceActor线程
    //停止 移除NN的BPOfferService线程
    //更新NN列表
}
BPServiceActor线程启动，进入BPServiceActor的run方法中，主要是与NN握手和发送心跳服务
public void run() {

    // 与namonode握手，注册
    connectToNNAndHandshake();

    // BPServiceActor提供的服务
    offerService();

}

进入connectToNNAndHandshake()，获取NN代理/名称空间信息/

private void connectToNNAndHandshake() throws IOException {
    // get NN proxy
    bpNamenode = dn.connectToNN(nnAddr);

    // First phase of the handshake with NN - get the namespace
    // info.
    NamespaceInfo nsInfo = retrieveNamespaceInfo();

    // Verify that this matches the other NN in this HA pair.
    // This also initializes our block pool in the DN if we are
    // the first NN connection for this BP.
    //验证/设置名称空间
    bpos.verifyAndSetNamespaceInfo(this, nsInfo);

    /* set thread name again to include NamespaceInfo when it's available. */
    this.bpThread.setName(formatThreadName("heartbeating", nnAddr));

    // Second phase of the handshake with the NN.
    //注册
    register(nsInfo);
  }

进入verifyAndSetNamespaceInfo(this, nsInfo)，设置名称空间，初始化块池

void verifyAndSetNamespaceInfo(BPServiceActor actor, NamespaceInfo nsInfo)
    throws IOException {
    try {
      DataNodeFaultInjector.get().delayWhenOfferServiceHoldLock();
      if (setNamespaceInfo(nsInfo) == null) {
        boolean success = false;

        // Now that we know the namespace ID, etc, we can pass this to the DN.
        // The DN can now initialize its local storage if we are the
        // first BP to handshake, etc.
        try {
          // 初始化块池对应的DataStorage、初始化FsDatasetImpl对象；
          // 初始化启动DataBlockScanner数据块扫描线程、和DirectoryScanners目录检测线程
          dn.initBlockPool(this);
          success = true;
        } finally {
          if (!success) {
            // The datanode failed to initialize the BP. We need to reset
            // the namespace info so that other BPService actors still have
            // a chance to set it, and re-initialize the datanode.
            setNamespaceInfo(null);
          }
        }
      }
    } finally {
      writeUnlock();
    }
  }