（一）、环境介绍
1、服务器的相关设置：
1、kafka1:172.20.67.52
2、kafka2:172.20.67.56
3、kafka3:172.20.67.57
2、zookeeper所需的三个端口
1、2181
2、2888
3、3888
三个端口的作用：
2181:对客户端提供服务
2888:集群内机器通讯使用（Leader监听此端口）
3888:选举leader使用
3、kafka 所需的端口
9092：kafka所需的端口
（二）、kafka集群的安装配置（1）、kafka1172.20.67.52的安装及相关配置
1、去官网下载相应的版本（https://kafka.apache.org/downloads )。我已下载好直接使用kafka_2.11-0.11.0.3版本
2、解压并放到相应的目录
[[email protected] opt]# tar xf kafka_2.11-0.11.0.3.tgz
[[email protected] opt]# mv kafka_2.11-0.11.0.3 /usr/local/kafka
3、zookeeper的相关配置[[email protected] kafka]# vim /usr/local/kafka/config/zookeeper.properties
dataDir=/data/zookeeper
clientPort=2181
maxClientCnxns=0
tickTime=2000
initLimit=20
syncLimit=10
server.1=172.20.67.52:2888:3888
server.2=172.20.67.56:2888:3888
server.3=172.20.67.57:2888:3888
4、kafka的相关配置
[[email protected] kafka]# vim /usr/local/kafka/config/server.properties
broker.id=1
host.name=172.20.67.52
listeners=PLAINTEXT://172.20.67.52:9092
log.dirs=/data/kafka-logs
log.retention.check.interval.ms=300000
log.retention.hours=1
log.segment.bytes=1073741824
num.io.threads=4
default.replication.factor=2
num.network.threads=3
num.partitions=30
num.recovery.threads.per.data.dir=1
port=9092
socket.receive.buffer.bytes=102400
socket.request.max.bytes=104857600
socket.send.buffer.bytes=102400
zookeeper.connect=172.20.67.52:2181,172.20.67.56:2181,172.20.67.57:2181
zookeeper.connection.timeout.ms=10000
fetch.message.max.bytes=52428700
replica.fetch.max.bytes=524287005、启动时加入到/etc/rc.local
[[email protected] config]# vim /etc/rc.local
cd /var/tmp;nohup /usr/local/kafka/bin/zookeeper-server-start.sh /usr/local/kafka/config/zookeeper.properties &
cd /tmp;nohup /usr/local/kafka/bin/kafka-server-start.sh /usr/local/kafka/config/server.properties &
touch /var/lock/subsys/local6、依次创建zookeeper的数据目录/data/zookeeper和文件/data/zookeeper/myid,此时我们需要对应的Zookeeper服务器上在dataDir的目录下面新建myid文件，同时把对应的1、2、3写进去。
[[email protected] config]# mkdir -p /data/zookeeper
[[email protected] config]# vim /data/zookeeper/myid
1
7、依次启动集群的kafka1、kafka2、kafka3 的zookeeper [[email protected] config]# /usr/local/kafka/bin/zookeeper-server-start.sh /usr/local/kafka/config/zookeeper.properties &
8、然后在依次启动集群的kafka1、kafka2、kafka3 的kafka服务[[email protected] config]# cd /tmp;nohup /usr/local/kafka/bin/kafka-server-start.sh /usr/local/kafka/config/server.properties &
9、最后依次按照2、3、4、5、6、7、8的步骤依次在kafka2（172.20.67.56）和kafka3（172.20.67.57）依次安装配置zookeeper和kafka的配置文件和启动。
10、查看集群所启动的端口
kafka1
[[email protected] ~]# netstat -lntp|grep java
tcp6 0 0 172.20.67.52:3888 :::* LISTEN 20184/java
tcp6 0 0 127.0.0.1:9600 :::* LISTEN 20143/java
tcp6 0 0 172.20.67.52:9092 :::* LISTEN 20649/java
tcp6 0 0 :::35300 :::* LISTEN 20649/java
tcp6 0 0 :::2181 :::* LISTEN 20184/java
tcp6 0 0 :::46471 :::* LISTEN 20184/java
kafka2
[[email protected] ~]# netstat -lntp|grep java
tcp6 0 0 172.20.67.56:3888 :::* LISTEN 30298/java
tcp6 0 0 :::33395 :::* LISTEN 30298/java
tcp6 0 0 :::43031 :::* LISTEN 30573/java
tcp6 0 0 172.20.67.56:9092 :::* LISTEN 30573/java
tcp6 0 0 :::2181 :::* LISTEN 30298/java
kafka3
[[email protected] ~]# netstat -lntp|grep java
tcp6 0 0 172.20.67.57:3888 :::* LISTEN 26313/java
tcp6 0 0 :::44194 :::* LISTEN 26313/java
tcp6 0 0 172.20.67.57:9092 :::* LISTEN 26883/java
tcp6 0 0 :::2181 :::* LISTEN 26313/java
tcp6 0 0 172.20.67.57:2888 :::* LISTEN 26313/java
tcp6 0 0 :::36041 :::* LISTEN 26883/java 原文地址：http://blog.51cto.com/liqingbiao/2315959