1、簡要介紹

　　zookeeper是一個分布式的應(yīng)用程序協(xié)調(diào)服務(wù)，是Hadoop和Hbase的重要組件，是一個樹型的目錄服務(wù)，支持變更推送。除此還可以用作dubbo服務(wù)的注冊中心。

2、安裝

　　2.1 下載安裝

wget http://mirrors.cnnic.cn/apache/zookeeper/zookeeper-3.4.6/zookeeper-3.4.6.tar.gztar -zxvf zookeeper-3.4.6.tar.gzcd zookeeper-3.4.6cp conf/zoo_sample.cfg conf/zoo.cfg

　　2.2 配置

　　　　2.2.1 單點方式

　　　　（1）修改zoo.cfg，如果沒有特殊要求，全部默認也可以，主要修改的地方就是dataDir 和 clientPort，如下：

tickTime=2000initLimit=10syncLimit=5dataDir=/app/soft/zookeeper-3.4.6/data （換成真實輸出目錄）clientPort=2181#tickTime：這個時間是作為 Zookeeper 服務(wù)器之間或客戶端與服務(wù)器之間維持心跳的時間間隔，也就是每個 tickTime 時間就會發(fā)送一個心跳，以毫秒為單位。#initLimit：LF初始通信時限，集群中的follower服務(wù)器(F)與leader服務(wù)器(L)之間初始連接時能容忍的最多心跳數(shù)（tickTime的數(shù)量）#syncLimit：集群中的follower服務(wù)器與leader服務(wù)器之間請求和應(yīng)答之間能容忍的最多心跳數(shù)（tickTime的數(shù)量）。#dataDir：顧名思義就是 Zookeeper 保存數(shù)據(jù)的目錄，默認情況下，Zookeeper 將寫數(shù)據(jù)的日志文件也保存在這個目錄里。#clientPort：這個端口就是客戶端連接 Zookeeper 服務(wù)器的端口，Zookeeper 會監(jiān)聽這個端口，接受客戶端的訪問請求。　　　#dataLogDir：日志文件目錄，Zookeeper保存日志文件的目錄#服務(wù)器名稱與地址：集群信息（服務(wù)器編號，服務(wù)器地址，LF通信端口，選舉端口），規(guī)則如：server.N=yyy:A:B#其中N表示服務(wù)器編號，YYY表示服務(wù)器的ip地址，A為LF通信端口，表示該服務(wù)器與集群中的leader交換的信息的端口。B為選舉端口，表示選舉新leader時服務(wù)器間相互通信的端口（當leader掛掉時，其余服務(wù)器會相互通信，選擇出新的leader）。一般來說，集群中每個服務(wù)器的A端口都是一樣，每個服務(wù)器的B端口也是一樣。但是當所采用的為偽集群時，IP地址都一樣，只能時A端口和B端口不一樣。

　　　　（2）啟動：bin/zkServer.sh start

　　　　（3）查看是否成功：bin/zkServer.sh status

　　　　（4）查看日志：vi zooKeeper.out

　　　　2.2.2 集群方式（單IP多節(jié)點）

　　　　（1）拷貝3份zookeeper-3.4.6.tar.gz，如zookeeper_node1、zookeeper_node2、zookeeper_node3，結(jié)構(gòu)如下：

　　　　　　　　　　（2）進入zookeeper_node1--->conf，修改zoo.cfg，如下：

tickTime=2000initLimit=10syncLimit=5dataDir=/app/soft/zookeeper_node1/data （換成真實輸出目錄）clientPort=2181server.1=127.0.0.1:2888:3888server.2=127.0.0.1:2889:3889server.3=127.0.0.1:2890:3890

　　　　（3）然后在上面dataDir對應(yīng)的目錄下創(chuàng)建myid文件，如下：

mkdir datavi myid

　　　　myid指明自己的id，對應(yīng)上面zoo.cfg中"server."后的數(shù)字，第一臺的內(nèi)容為1，第二臺的內(nèi)容為2，第三臺的內(nèi)容為3，內(nèi)容如下：

1

　　　　（4）依次類推，調(diào)整2、3節(jié)點的地址及端口，如下：

節(jié)點2：　　　　　　

tickTime=2000initLimit=10syncLimit=5dataDir=/app/soft/zookeeper_node2/data （換成真實輸出目錄）clientPort=2182server.1=127.0.0.1:2888:3888server.2=127.0.0.1:2889:3889server.3=127.0.0.1:2890:3890

節(jié)點3：

tickTime=2000initLimit=10syncLimit=5dataDir=/app/soft/zookeeper_node3/data （換成真實輸出目錄）clientPort=2183server.1=127.0.0.1:2888:3888server.2=127.0.0.1:2889:3889server.3=127.0.0.1:2890:389

　　　　（5），修改節(jié)點2、3的myid；　　　　　　

　　　　（6）啟動服務(wù)，如下：　　

[root@localhost soft]# zookeeper_node1/bin/zkServer.sh startJMX enabled by defaultUsing config: /app/soft/zookeeper_node1/bin/../conf/zoo.cfgStarting zookeeper ... STARTED[root@localhost soft]# vi zookeeper.out 2015-06-25 05:43:13,252 [myid:] - INFO  [main:QuorumPeerConfig@103] - Reading configuration from: /app/soft/zookeeper_nod2015-06-25 05:43:13,257 [myid:] - INFO  [main:QuorumPeerConfig@340] - Defaulting to majority quorums2015-06-25 05:43:13,260 [myid:1] - INFO  [main:DatadirCleanupManager@78] - autopurge.snaPRetainCount set to 32015-06-25 05:43:13,260 [myid:1] - INFO  [main:DatadirCleanupManager@79] - autopurge.purgeInterval set to 02015-06-25 05:43:13,262 [myid:1] - INFO  [main:DatadirCleanupManager@101] - Purge task is not scheduled.2015-06-25 05:43:13,273 [myid:1] - INFO  [main:QuorumPeerMain@127] - Starting quorum peer2015-06-25 05:43:13,285 [myid:1] - INFO  [main:NIOServerCnxnFactory@94] - binding to port 0.0.0.0/0.0.0.0:21812015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@959] - tickTime set to 20002015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@979] - minsessionTimeout set to -12015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@990] - maxSessionTimeout set to -12015-06-25 05:43:13,315 [myid:1] - INFO  [main:QuorumPeer@1005] - initLimit set to 102015-06-25 05:43:13,359 [myid:1] - INFO  [Thread-1:QuorumCnxManager$Listener@504] - My election bind port: /127.0.0.1:3882015-06-25 05:43:13,371 [myid:1] - INFO  [QuorumPeer[myid=1]/0:0:0:0:0:0:0:0:2181:QuorumPeer@714] - LOOKING2015-06-25 05:43:13,374 [myid:1] - INFO  [QuorumPeer[myid=1]/0:0:0:0:0:0:0:0:2181:FastLeaderElection@815] - New election.2015-06-25 05:43:13,252 [myid:] - INFO  [main:QuorumPeerConfig@103] - Reading configuration from: /app/soft/zookeeper_node1/bin/../conf/zoo.cfg2015-06-25 05:43:13,257 [myid:] - INFO  [main:QuorumPeerConfig@340] - Defaulting to majority quorums2015-06-25 05:43:13,260 [myid:1] - INFO  [main:DatadirCleanupManager@78] - autopurge.snapRetainCount set to 32015-06-25 05:43:13,260 [myid:1] - INFO  [main:DatadirCleanupManager@79] - autopurge.purgeInterval set to 02015-06-25 05:43:13,262 [myid:1] - INFO  [main:DatadirCleanupManager@101] - Purge task is not scheduled.2015-06-25 05:43:13,273 [myid:1] - INFO  [main:QuorumPeerMain@127] - Starting quorum peer2015-06-25 05:43:13,285 [myid:1] - INFO  [main:NIOServerCnxnFactory@94] - binding to port 0.0.0.0/0.0.0.0:21812015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@959] - tickTime set to 20002015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@979] - minSessionTimeout set to -12015-06-25 05:43:13,312 [myid:1] - INFO  [main:QuorumPeer@990] - maxSessionTimeout set to -12015-06-25 05:43:13,315 [myid:1] - INFO  [main:QuorumPeer@1005] - initLimit set to 102015-06-25 05:43:13,359 [myid:1] - INFO  [Thread-1:QuorumCnxManager$Listener@504] - My election bind port: /127.0.0.1:38882015-06-25 05:43:13,371 [myid:1] - INFO  [QuorumPeer[myid=1]/0:0:0:0:0:0:0:0:2181:QuorumPeer@714] - LOOKING2015-06-25 05:43:13,374 [myid:1] - INFO  [QuorumPeer[myid=1]/0:0:0:0:0:0:0:0:2181:FastLeaderElection@815] - New election. My id =  1, proposed zxid=0x22015-06-25 05:43:13,376 [myid:1] - INFO  [WorkerReceiver[myid=1]:FastLeaderElection@597] - Notification: 1 (message format version), 1 (n.leader), 0x2 (n.zxid), 0x1 (n.round), LOOKING (n.state), 1 (n.sid), 0x1 (n.peerEpoch) LOOKING (my state)2015-06-25 05:43:13,379 [myid:1] - WARN  [WorkerSender[myid=1]:QuorumCnxManager@382] - Cannot open channel to 2 at election address /127.0.0.1:3889java.net.ConnectException: Connection refused        at java.net.PlainSocketImpl.socketConnect(Native Method)        at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:339)        at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:200)        at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:182)        at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:392)        at java.net.Socket.connect(Socket.java:579)        at org.apache.zookeeper.server.quorum.QuorumCnxManager.connectOne(QuorumCnxManager.java:368)        at org.apache.zookeeper.server.quorum.QuorumCnxManager.toSend(QuorumCnxManager.java:341)        at org.apache.zookeeper.server.quorum.FastLeaderElection$Messenger$WorkerSender.process(FastLeaderElection.java:449)        at org.apache.zookeeper.server.quorum.FastLeaderElection$Messenger$WorkerSender.run(FastLeaderElection.java:430)        at java.lang.Thread.run(Thread.java:745)2015-06-25 05:43:13,385 [myid:1] - WARN  [WorkerSender[myid=1]:QuorumCnxManager@382] - Cannot open channel to 3 at election address /127.0.0.1:3890java.net.ConnectException: Connection refused        at java.net.PlainSocketImpl.socketConnect(Native Method)        at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:339)        at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:200)

　　　　發(fā)現(xiàn)報錯，是因為2、3節(jié)點服務(wù)還未啟動。按照節(jié)點1的啟動方式，依次啟動2、3節(jié)點，再次查看日志，發(fā)現(xiàn)服務(wù)正常。

　　　　查看日志命令：vi zookeeper.out

　　　　（7）查看服務(wù)狀態(tài)，如下：　　　　

[root@localhost soft]# zookeeper_node1/bin/zkServer.sh statusJMX enabled by defaultUsing config: /app/soft/zookeeper_node1/bin/../conf/zoo.cfgMode: follower[root@localhost soft]# zookeeper_node2/bin/zkServer.sh statusJMX enabled by defaultUsing config: /app/soft/zookeeper_node2/bin/../conf/zoo.cfgMode: leader[root@localhost soft]# zookeeper_node3/bin/zkServer.sh statusJMX enabled by defaultUsing config: /app/soft/zookeeper_node3/bin/../conf/zoo.cfgMode: follower

　　　　　　可以看出1是follower，2是leader，3是follower。