Red Hat Linux 6.5 安装Zookeeper集群

一，安装环境：

操作系统：Red Hat Linux 6.5

JDK版本：1.8.0_151 //需先安装JDK

Zookeeper版本：3.4.12

设置每个节点的hosts文件。

# vi /etc/hosts
//添加以下内容
10.11.111.42  mrzwk-5
10.11.111.43  mrzwk-6
10.11.111.45  mrzwk-8

重启生效。（reboot）

二、安装zookeeper集群

# cd /home/weihu
# mkdir kafka
//解压
# tar -zxvf zookeeper-3.4.12.tar.gz -C kafka     
#cd kafka
//创建链接
# ln -s zookeeper-3.4.12 zookeeper       
# cd zookeeper
//创建日志目录
# mkdir -pv logs
//创建数据目录
# mkdir -pv data

配置zookeeper环境变量

# vi /etc/profile
export PATH=/home/weihu/kafka/zookeeper/bin:$PATH

保存退出
//初始化
# source /etc/profile

配置zookeeper集群：

集群节点数量要为奇数，且至少有3个节点，因为ZooKeeper集群是以宕机个数过半才会让整个集群宕机的。搭建ZooKeeper集群时，一定要先停止所有已经启动的节点。

修改配置文件：

# cd /home/weihu/kafka/zookeeper/conf
# cp zoo_sample.cfg zoo.cfg
#vi zoo.cfg
增加或修改成如下代码：

# The number of milliseconds of each tick
tickTime=2000
# The number of ticks that the initial 
# synchronization phase can take
initLimit=10
# The number of ticks that can pass between 
# sending a request and getting an acknowledgement
syncLimit=5
# the directory where the snapshot is stored.
# do not use /tmp for storage, /tmp here is just 
# example sakes.
dataDir=/home/weihu/kafka/zookeeper/data
dataLogDir=/home/weihu/kafka/zookeeper/logs
# the port at which the clients will connect
clientPort=2181
server.1=10.11.111.42:2888:3888
server.2=10.11.111.43:2888:3888
server.3=10.11.111.45:2888:3888

# the maximum number of client connections.
# increase this if you need to handle more clients
maxClientCnxns=60

创建myid文件：

在mrzwk-5上执行：# echo 1 > /home/weihu/kafka/zookeeper/data/myid
在mrzwk-6上执行：# echo 2 > /home/weihu/kafka/zookeeper/data/myid
在mrzwk-8上执行：# echo 3 > /home/weihu/kafka/zookeeper/data/myid

备注：myid文件中只有一行内容，且内容为该节点对应的server.id中的id编号

启动zookeeper集群：

# cd /home/weihu/kafka/zookeeper/bin
# zkServer.sh start
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Starting zookeeper ... STARTED

查看zookeeper进程：

# jps  //显示当前所有Java进程的PID
6470 Jps
6316 QuorumPeerMain

查看zookeeper集群状态：

# zkServer.sh status

上述步骤分别在三台服务器上执行，

在mrzwk-5上：

# zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Mode: follower

在mrzwk-6上：

# zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Mode: follower

在mrzwk-8上：

# zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Mode: leader

备注：正常情况下应该是1个leader节点，2个follower节点

若zkServer.sh status命令执行显示-bash: zkServer.sh: command not found

命令改为./zkServer.sh status可正常执行

连接zookeeper集群：

# zkCli.sh -server mrzwk-5:2181,mrzwk-6:2181,mrzwk-8:2181

成功连接后，可以看到如下输出：

Welcome to ZooKeeper!
2018-05-04 10:39:27,990 [myid:] - INFO  [main-SendThread(mrzwk-5:2181):ClientCnxn$SendThread@1028] - Opening socket connection to server mrzwk-5/10.11.111.42:2181. Will not attempt to authenticate using SASL (unknown error)
JLine support is enabled
2018-05-04 10:39:28,071 [myid:] - INFO  [main-SendThread(mrzwk-5:2181):ClientCnxn$SendThread@878] - Socket connection established to mrzwk-5/10.11.111.42:2181, initiating session
[zk: mrzwk-5:2181,mrzwk-6:2181,mrzwk-8:2181(CONNECTING) 0] 2018-05-04 10:39:28,116 [myid:] - INFO  [main-SendThread(mrzwk-5:2181):ClientCnxn$SendThread@1302] - Session establishment complete on server mrzwk-5/10.11.111.42:2181, sessionid = 0x30000a829550000, negotiated timeout = 30000
WATCHER::
WatchedEvent state:SyncConnected type:None path:null
[zk: mrzwk-5:2181,mrzwk-6:2181,mrzwk-8:2181(CONNECTED) 0]

三，可能遇到的报错

1，报错:Zookeeper启动成功，zkServer.sh status 报错

# zkServer.sh start
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Starting zookeeper ... STARTED
# zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /home/weihu/kafka/zookeeper/bin/../conf/zoo.cfg
Error contacting service. It is probably not running.

原因：

1. zoo.cfg文件配置出错：dataLogDir指定的目录未被创建
2. myid文件中的整数格式不对，或者与zoo.cfg中的server整数不对应
3. 防火墙相关端口是否开启（2181,2888,3888）
4. ZooKeeper集群IP不能正确映射
   单机模式正常，且单机模式下能互相访问，切换至分布式Zookeeper启动成功，zkServer.sh status 报错。
   zookeeper.out日志输出如下：
   

   [myid:1] - ERROR [centos_1/220.250.64.225:3888:QuorumCnxManager$Listener@562] - As I 'm leaving the listener thread, I won' t be able to participate in leader election any longer: centos_1/220.250.64.225:38882016-11-03 19:33:43,795 [myid:1] - WARN  [WorkerSender[myid=1]:QuorumCnxManager@400] - Cannot open channel to 2 at election address centos_2/220.250.64.225:3888java.net.SocketTimeoutException: connect timed out
       at java.net.PlainSocketImpl.socketConnect(Native Method)

   
   解决：更改配置文件zoo.cfg
   在mrzwk-5上：
   

5. server.1=0.0.0.0:2888:3888
   server.2=10.11.111.43:2888:3888
   server.3=10.11.111.45:2888:3888

   在mrzwk-6上：
   

   server.1=10.11.111.42:2888:3888
   server.2=0.0.0.0:2888:3888
   server.3=10.11.111.45:2888:3888

   在mrzwk-8上：
   

   server.1=10.11.111.42:2888:3888
   server.2=10.11.111.43:2888:3888
   server.3=0.0.0.0:2888:3888

保存配置文件，并重启服务。