zookeeper学习系列：二、api实践

上一章我们知道zookeeper的简介，启动，设置节点以及结构性能。本小节我们来玩玩api，获取下数据。

php版本： http://anykoro.sinaapp.com/2013/04/05/%E4%BD%BF%E7%94%A8apache-zookeeper%E5%88%86%E5%B8%83%E5%BC%8F%E9%83%A8%E7%BD%B2php%E5%BA%94%E7%94%A8%E7%A8%8B%E5%BA%8F/

go版本：http://mmcgrana.github.io/2014/05/getting-started-with-zookeeper-and-go.html

读一下：http://zookeeper.apache.org/doc/trunk/javaExample.html

然后我说 what the fuck it is？

我就想读个数据，需要这么复杂么。。。

动手改一下

版本1:  只获取数据，不管别的：

import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.ZooKeeper;
import java.io.IOException;
public class ZkReader {
    public static void main(String[] args) throws IOException, InterruptedException, KeeperException {
        String hostPort = "192.168.1.2,192.168.1.3,192.168.1.4";
        String znode = "/test";
        ZooKeeper zk = new ZooKeeper(hostPort, 3000, null);
        System.out.println(new String(zk.getData(znode,false,null)));
    }
}

在zkcli上创建 /test 并改变它的值：123，运行，输出：

123

能得到结果，但是报错了：

14/10/17 11:51:58 ERROR zookeeper.ClientCnxn: Error while calling watcher
java.lang.NullPointerException
    at org.apache.zookeeper.ClientCnxn$EventThread.processEvent(ClientCnxn.java:521)
    at org.apache.zookeeper.ClientCnxn$EventThread.run(ClientCnxn.java:497)

看下源码，需要注册个watcher，意思是不这样zookeeper就只是个纯配置了？ok

版本2：zk get data＋watcher

import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooKeeper;
import java.io.IOException;

public class ZkReader {
    public static void main(String[] args) throws IOException, InterruptedException, KeeperException {
        String hostPort = "10.16.73.22,10.16.73.12,10.16.73.13";
        String znode = "/test";
        ZooKeeper zk = new ZooKeeper(hostPort, 3000, new MyWatcher());
        System.out.println(new String(zk.getData(znode,false,null)));
    }
}

class MyWatcher  implements Watcher {

    @Override
    public void process(WatchedEvent event) {
        System.out.println("hello zookeeper");
        System.out.println(String.format("hello event! type=%s, stat=%s, path=%s",event.getType(),event.getState(),event.getPath()));
    }
}

输出却是：

hello zookeeper
123
hello event! type=None, stat=SyncConnected, path=null

data总是在中间？百撕不得姐，在邮件组里咨询下，几天后有了回复（不够活跃的邮件组了）：

“

Zookeeper works asynchronously in several threads. Therefore the sequence of execution in different threads is not generally predictable. It could therefore happen that when the connection status change is detected, the Watcher is executed, but only the first "hello zookeeper" gets echoed, then the main thread gets some cycles again and prints "123", after which the second print statement "hello event!..." is executed. If you don't want this to happen, use a CountDownLatch to make the main thread wait until the Zookeeper connection is established and propertly recognized in your program. The main thread creates the CountDownLatch(1), opens the Zk connection and waits latch.await(). The Watcher does its job and then counts the latch down by one, causing the main thread to leave the await and continue doing its job.

”

被认为是多线程问题，建立zk连接时会启动多个线程：sendThread  eventThread

eventThread执行到一半时,主线程获得了cpu，打印出结果，然后eventThread继续执行watcher.process。

这两个版本只是做到了获取数据，如果数据有变动，需要自动更新呢？ok，参照zk给的例子，简化出第三个版本：

DataMonitor.java ：

/**
 * A simple class that monitors the data and existence of a ZooKeeper
 * node. It uses asynchronous ZooKeeper APIs.
 */
import java.util.Arrays;

import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooKeeper;
import org.apache.zookeeper.AsyncCallback.StatCallback;
import org.apache.zookeeper.data.Stat;

public class DataMonitor implements Watcher, StatCallback {

    ZooKeeper zk;

    String znode;

    boolean dead;

    DataMonitorListener listener;

    byte prevData[];

    public DataMonitor(ZooKeeper zk, String znode,  DataMonitorListener listener) {
        this.zk = zk;
        this.znode = znode;
        this.listener = listener;
        // Get things started by checking if the node exists. We are going
        // to be completely event driven
        zk.exists(znode, true, this, null);
    }

    /**
     * Other classes use the DataMonitor by implementing this method
     */
    public interface DataMonitorListener {
        /**
         * The existence status of the node has changed.
         */
        void showData(byte data[]);
    }

    public void process(WatchedEvent event) {
        String path = event.getPath();
        if (event.getType() != Event.EventType.None) {
            System.out.println("watch event type: "+event.getType());
            if (path != null && path.equals(znode)) {
                // Something has changed on the node, let's find out
                zk.exists(znode, true, this, null);
            }
        }
    }

    public void processResult(int rc, String path, Object ctx, Stat stat) {
        System.out.println("rc : "+rc);

        byte b[] = null;
            try {
                b = zk.getData(znode, false, null);
            } catch (KeeperException e) {
                // We don't need to worry about recovering now. The watch
                // callbacks will kick off any exception handling
                e.printStackTrace();
            } catch (InterruptedException e) {
                return;
            }
        if ((b == null && b != prevData)
                || (b != null && !Arrays.equals(prevData, b))) {
            listener.showData(b);
            prevData = b;
        }
    }
}

Executor.java：

import java.io.IOException;
import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooKeeper;
public class Executor
        implements Watcher, Runnable, DataMonitor.DataMonitorListener
{
    DataMonitor dm;
    ZooKeeper zk;

    public Executor(String hostPort, String znode) throws KeeperException, IOException {
        zk = new ZooKeeper(hostPort, 3000, this);
        dm = new DataMonitor(zk, znode, this);
    }

    public static void main(String[] args) {
        String hostPort = "192.168.1.22,192.168.1.12,192.168.1.13";
        String znode = "/test";
        try {
            new Executor(hostPort, znode).run();
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

    /***************************************************************************
     * We do process any events ourselves, we just need to forward them on.
     *
     * @see org.apache.zookeeper.Watcher#
     */
    public void process(WatchedEvent event) {
        System.out.println("Executor process event: "+event.getType());
        dm.process(event);

    }

    public void run() {
        try {
            synchronized (this) {
                while (true) {
                    wait();
                }
            }
        } catch (InterruptedException e) {
        }
    }

    public void showData(byte[] data) {
            System.out.println("data changes: "+new String(data));
    }
}

一个执行者一个监控，注册watcher到zk，当有事件发生时，推送本身的StatCallback到Zookeeper，当节点有变动时调用processResult展示结果。

Executor process event: NodeDataChanged
watch event type: NodeDataChanged
rc : 0
data changes: abcd

还是有点复杂，仔细看下DataMonitor似乎没有存在的必要，我只需要一个类，启动zk client，并监听数据变化就好了，于是有了第四个单对象版本：

Executor.java

import java.io.IOException;
import java.util.Arrays;

import org.apache.zookeeper.*;
import org.apache.zookeeper.data.Stat;

public class Executor
        implements Watcher, Runnable, AsyncCallback.StatCallback
{
    ZooKeeper zk;
    String znode;
    byte prevData[];

    public Executor(String hostPort, String znode) throws KeeperException, IOException {
        zk = new ZooKeeper(hostPort, 3000, this);
        this.znode = znode;
        zk.exists(znode, true, this, null);
    }

    public static void main(String[] args) {
        String hostPort = "10.16.73.22,10.16.73.12,10.16.73.13";
        String znode = "/test";
        try {
            new Executor(hostPort, znode).run();
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

    /***************************************************************************
     * We do process any events ourselves, we just need to forward them on.
     *
     * @see org.apache.zookeeper.Watcher#
     */
    public void process(WatchedEvent event) {
        String path = event.getPath();
        if (event.getType() != Event.EventType.None) {
            System.out.println("watch event type: "+event.getType());
            if (path != null && path.equals(znode)) {
                // Something has changed on the node, let's find out
                zk.exists(znode, true, this, null);
            }
        }

    }

    public void run() {
        try {
            synchronized (this) {
                while (true) {
                    wait();
                }
            }
        } catch (InterruptedException e) {
        }
    }

    public void processResult(int rc, String path, Object ctx, Stat stat) {
        System.out.println("rc : "+rc);

        byte b[] = null;
        try {
            b = zk.getData(znode, false, null);
        } catch (KeeperException e) {
            // We don't need to worry about recovering now. The watch
            // callbacks will kick off any exception handling
            e.printStackTrace();
        } catch (InterruptedException e) {
            return;
        }
        if ((b == null && b != prevData)
                || (b != null && !Arrays.equals(prevData, b))) {
            System.out.println("data changes: "+new String(b));
            prevData = b;
        }
    }
}

自己做watcher，并注册回调函数给zk，更简洁。

经测试，zk三台停掉一台，剩一主一从，仍能正常服务，剩一台时则报错，无法连接，重启动zk变成两台，客户端也无法恢复，重启了才恢复。

看了php api，理解了一下zk.exists 做的操作，exists和get方法都会注册回调过去，一个是注册watcher，一个是注册StatCallback，当触发事件时，监视器会被消费掉，所以我们需要在回调函数中再次设置监视器。于是有了第五个版本

import java.io.IOException;

import org.apache.zookeeper.*;

public class Executor
        implements Watcher, Runnable
{
    ZooKeeper zk;
    String znode;

    public Executor(String hostPort, String znode) throws KeeperException, IOException, InterruptedException {
        zk = new ZooKeeper(hostPort, 30000, this);
        this.znode = znode;
        zk.getData(znode, this, null);
    }

    public static void main(String[] args) {
        String hostPort = "10.16.73.22,10.16.73.12,10.16.73.13";
        String znode = "/test";
        try {
            new Executor(hostPort, znode).run();
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

    public void process(WatchedEvent event) {
        String path = event.getPath();
        if (event.getType() != Event.EventType.None) {
            System.out.println("watch event type: "+event.getType());
            if (path != null && path.equals(znode)) {
                // Something has changed on the node, let's find out
                try {
                    System.out.println(new String(zk.getData(znode, this, null)));
                } catch (KeeperException e) {
                    e.printStackTrace();  //To change body of catch statement use File | Settings | File Templates.
                } catch (InterruptedException e) {
                    e.printStackTrace();  //To change body of catch statement use File | Settings | File Templates.
                }
            }
        }

    }

    public void run() {
        try {
            synchronized (this) {
                while (true) {
                    wait();
                }
            }
        } catch (InterruptedException e) {
        }
    }
}

上边这两个版本已经可以检测到zk的数据节点变动，但没有处理异常情况，没有处理close事件，大家可以自己动手改造下难懂的http://zookeeper.apache.org/doc/trunk/javaExample.html  例子。

更多java api操作（创建节点、删除修改等）：http://www.cnblogs.com/haippy/archive/2012/07/19/2600032.html