zookeeper的Java客户端API
zookeeper作为一个分布式服务框架,主要用来解决分布式数据一致性问题,对多种语言提供了API。这里主要记录下JAVA客户端API的使用。
1.创建会话
客户端可以通过创建一个ZooKeeper实例来连接zookeeper服务器
ZooKeeper的4个构造函数如下:
- ZooKeeper(connectString, sessionTimeout, watcher);
- ZooKeeper(connectString, sessionTimeout, watcher,canBeReadOnly);
- ZooKeeper(connectString, sessionTimeout, watcher, sessionId, sessionPasswd);
- ZooKeeper(connectString, sessionTimeout, watcher, sessionId, sessionPasswd, canBeReadOnly);
参数说明:
connectString: 连接字符串 例如 "127.0.0.1:2181"
sessionTimeout: 会话超时时间 以毫秒为单位的整型值 在sessionTimeout时间内服务端与客户端没有有效的心跳检测 则会话失效
watcher: 默认的事件通知处理器
sessionId: 会话ID
sessionPasswd: 会话秘钥
canBeReadOnly: 是否是只读
- public class ZookeeperSampleTest implements Watcher {
- private static CountDownLatch connectedSemaphore = new CountDownLatch(1);
- public static void main(String args[]) throws IOException, InterruptedException{
- ZooKeeper zk = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleTest());
- System.out.println(zk.getState());
- try {
- connectedSemaphore.await();
- } catch (Exception e) {
- // TODO: handle exception
- }
- Long sessionId = zk.getSessionId();
- byte[] password = zk.getSessionPasswd();
- zk = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleTest(),1L,"test".getBytes());
- zk = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleTest(),sessionId,password);
- Thread.sleep(Integer.MAX_VALUE);
- //System.out.println("ZooKeeper session established");
- }
- @Override
- public void process(WatchedEvent event) {
- // TODO Auto-generated method stub
- System.out.println("Receive watched event :" + event);
- if(event.getState()==KeeperState.SyncConnected){
- connectedSemaphore.countDown();
- }
- }
- }
2.创建节点
创建节点的方式分为同步和异步,构造函数如下:
- zk.create(path, data, acl, createMode);
- , data, acl, createMode, cb, ctx);
参数说明:
path: 创建节点的路径 如:/zk-test
data[]: 字节数组,节点数据
acl: 权限
createMode: 节点类型 分为4种: 持久性节点,持久性有序节点,临时节点,临时有序节点
cb: 回调函数 需要实现StringCallback接口
ctx: 上下文 一般用于回调函数时候使用
- public class ZookeeperSampleCreateTest implements Watcher {
- public static CountDownLatch connectedSemaphore = new CountDownLatch(1);
- public static void main(String[] args) throws IOException, InterruptedException, KeeperException{
- ZooKeeper zk = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleCreateTest());
- connectedSemaphore.await();
- /*同步的方式*/
- /*String path1 = zk.create("/zk-test-ephemeral-", "test-ephemeral".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL);
- System.out.println("Success create node :" + path1);
- String path2 = zk.create("/zk-test-ephemeral-", "test-ephemeral-sequential".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL_SEQUENTIAL);
- System.out.println("Success create node :" + path2);*/
- /*异步的方式*/
- zk.create("/zk-test-ephemeral-", "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL,new IStringCallback(),"i am context");
- zk.create("/zk-test-ephemeral-", "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL,new IStringCallback(),"i am context");
- zk.create("/zk-test-ephemeral-", "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL_SEQUENTIAL,new IStringCallback(),"i am context");
- Thread.sleep(Integer.MAX_VALUE);
- }
- @Override
- public void process(WatchedEvent event) {
- // TODO Auto-generated method stub
- if(KeeperState.SyncConnected == event.getState()){
- connectedSemaphore.countDown();
- }
- }
- }
- class IStringCallback implements AsyncCallback.StringCallback{
- @Override
- public void processResult(int rc, String path, Object ctx, String name) {
- // TODO Auto-generated method stub
- System.out.println("Create path result: [" + rc + "," + path + "," + ctx + ", real path name " + name);
- }
- }
3.删除节点
- zk.delete(path,version);
- zk.delete(path,version,cb,ctx);
参数说明:
path:节点路径
version:版本号
cb: 回调函数
ctx: 上下文
4.获取子节点
getChildren有8个重载方法
- zk.getChildren(path, watch);
- zk.getChildren(path, watcher);
- zk.getChildren(path, watch, stat);
- zk.getChildren(path, watcher, stat);
- zk.getChildren(path, watch, cb, ctx);
- zk.getChildren(path, watcher, cb, ctx);
其中 回调函数有两种 ChildrenCallback Children2Callback
参数说明:
path:节点路径
watch: boolean类型 如果为true 表示使用默认的Watcher 为false表示不需要Watcher
watcher: 通知处理器 在本次获取子节点以后 一旦子节点有变化机会收到服务端传来的通知
stat: 指定数据节点的节点状态信息,传入一个旧的stat对象,当执行方法后 stat会被服务器响应的新stat替换
cb:回调函数 有两种类型 上面已经说过
ctx: 上下文
5.获取节点数据
获取节点数据有4个重载方法
- zk.getData(path, watch, stat);
- zk.getData(path, watcher, stat);
- zk.getData(path, watch, cb, ctx);
- zk.getData(path, watcher, cb, ctx);
参数说明:
path: 节点路径
watch: boolean类型 如果为true 表示使用默认的Watcher 为false表示不需要Watcher
stat: 指定数据节点的节点状态信息,传入一个旧的stat对象,当执行方法后 stat会被服务器响应的新stat替换
cb:回调函数 有两种类型 上面已经说过
ctx: 上下文
在获取节点数据时候 如果注册watcher 在节点数据发送变化的时候会通知客户端,当客户端收到通知以后,如果想下次数据发送变化再次收到通知,
需要重新注册watcher,获取子节点机制也如此
6.更新节点数据
更新节点数据也分为同步异步两个方法
- zk.setData(path, data, version);
- zk.setData(path, data, version, cb, ctx);
参数说明:同上
- public class ZookeeperSampleGetDataTest implements Watcher {
- public static CountDownLatch connectedSemaphore = new CountDownLatch(1);
- public static Stat stat = new Stat();
- public static ZooKeeper zk = null;
- public static void main(String[] args) throws IOException, KeeperException, InterruptedException{
- String path = "/zk-test";
- zk = new ZooKeeper("192.168.1.138:2181", 5000, new ZookeeperSampleGetDataTest());
- connectedSemaphore.await();
- //获取子节点
- /*zk.delete(path, 0);
- zk.create(path, "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);
- zk.create(path+"/c1", "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL);
- //同步方法获取children
- //List<String> childs = zk.getChildren(path, true);
- //System.out.println(childs);
- //异步方法获取children
- zk.getChildren(path, true, new IChildren2Callback(), null);
- zk.create(path+"/c2", "".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL);
- Thread.sleep(Integer.MAX_VALUE);*/
- //获取节点数据
- zk.delete(path, 0);
- zk.create(path, "123".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL);
- System.out.println(new String(zk.getData(path, true, stat)));
- /** czxid: 创建该节点的事务ID mzxid: 更新该节点的事务ID version: 数据版本*/
- System.out.println("Czxid: " + stat.getCzxid() + "Mzxid: " + stat.getMzxid() + "Version: " + stat.getVersion());
- zk.setData(path, "123".getBytes(), -1);
- Thread.sleep(Integer.MAX_VALUE);
- }
- @Override
- public void process(WatchedEvent event) {
- // TODO Auto-generated method stub
- if(KeeperState.SyncConnected==event.getState()){
- if(EventType.None==event.getType() && null==event.getPath()){
- connectedSemaphore.countDown();
- }else if(event.getType()==EventType.NodeChildrenChanged){
- try {
- System.out.println("Get Child :" + zk.getChildren(event.getPath(), true));
- } catch (KeeperException | InterruptedException e) {
- // TODO Auto-generated catch block
- e.printStackTrace();
- }
- }else if(event.getType()==EventType.NodeDataChanged){
- try {
- System.out.println(new String(zk.getData(event.getPath(), true, stat)));
- System.out.println("Czxid: " + stat.getCzxid() + "Mzxid: " + stat.getMzxid() + "Version: " + stat.getVersion());
- } catch (KeeperException | InterruptedException e) {
- // TODO Auto-generated catch block
- e.printStackTrace();
- }
- }
- }
- }
- }
- class IChildren2Callback implements AsyncCallback.Children2Callback{
- @Override
- public void processResult(int rc, String path, Object ctx,
- List<String> childrens, Stat stat) {
- // TODO Auto-generated method stub
- System.out.println("Get Children znode result: [response code: " + rc + ", param path: " + path + ", ctx " + ctx
- + ", childrens :" + childrens + ", stat: " + stat);
- }
- }
其中更新节点数据 版本version问题 -1表示基于数据的最新版本更新 这里可以作为分布式锁的一个思路 如果客户端参入的version不是数据最新版本则会更新失败
比如目前节点"/zk-test"的数据版本为 2 而某个客户端尝试 执行 setData("/zk-test","test".getBytes(),1) 由于传入version为1 < 服务器目前版本2 这样就会更新失败
7.检测节点是否存在
- zk.exists(path, watch);
- zk.exists(path, watcher);
- zk.exists(path, watch, cb, ctx);
- zk.exists(path, watcher, cb, ctx);
如果判断节点是否存在是 注册watcher 会对节点是否存在进行监听--创建节点,删除节点,节点数据更新都会通知客户端
8.权限控制
zookeeper提供了ACL的权限控制机制,简单来说就是通过控制zookeeper服务器上数据节点的ACL,来控制客户端对节点的访问权限
- addAuthInfo(String scheme,byte[] auth);
参数说明:
scheme: 权限控制模式 分为: world ,auth,digest,ip和super
auth: 具体的权限信息 类似于shiro的权限字符串
如下代码:
- <span style="white-space:pre;"> </span>ZooKeeper zk1 = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleCreateTest());
- zk1.addAuthInfo("digest", "test:true".getBytes());
- zk1.create("/zk-test-auth", "123".getBytes(), Ids.CREATOR_ALL_ACL, CreateMode.EPHEMERAL);
- ZooKeeper zk2 = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleCreateTest());
- zk2.addAuthInfo("digest", "test:true".getBytes());
- System.out.println(new String(zk2.getData("/zk-test-auth", false, null)));
- ZooKeeper zk3 = new ZooKeeper("192.168.1.138:2181",5000,new ZookeeperSampleCreateTest());
- zk3.addAuthInfo("digest", "test:false".getBytes());
- zk3.getData("/zk-test-auth", false, null);
zk2设置了正确的权限 所以可以获取到节点数据 zk3则会抛异常
org.apache.zookeeper.KeeperException$NoAuthException: KeeperErrorCode = NoAuth for /zk-test-auth