This is a record and then build a cluster, start the installation configuration zookeeper, its role is to make the cluster information synchronization, when the zookeeper configure itself a separate small clusters, cluster machine is generally an odd number, more than half as long as the machines work properly then the zookeeper cluster able to work, work automatically to elect a leader for the rest of the follower, so the minimum configuration is three.

Note that this article is not marked as almost all of the operations, it defaults to the user operating hadoop

1. First of all modifications on a write batch script

Copy, delete and then name the two machines within ips ,, leaving only the first three can be, then the name of several scripts are what change it, and the internal reference name will change it.

I can change a good https://www.lanzous.com/b849762/ Password: 1qq6

2. Install the zookeeper

Can be uploaded with xshell the rz command zookeeper installation package, the installation package here https://www.lanzous.com/b849708/ Password: 8a10

[hadoop@nn1 ~]$ cd zk_op/
批量发送给三台机器
[hadoop@nn1 zk_op]$ ./zk_scp_all.sh ~/upload/zookeeper-3.4.8.tar.gz /tmp/   
查看是否上传成功
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh "ls -l /tmp | grep zoo*"  
批量解压到各自的/usr/local/目录下
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh tar -zxf /tmp/zookeeper-3.4.8.tar.gz -C /usr/local/
再次查看是否操作成功
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh "ls -l /usr/local/ | grep zoo*"  
批量改变/usr/local/zookeeper-3.4.8目录的用户组为hadoop:hadoop
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh chown -R hadoop:hadoop /usr/local/zookeeper-3.4.8
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh chmod -R 770 /usr/local/zookeeper-3.4.8
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh "ls -l /usr/local/ | grep zookeeper-3.4.8"  
批量创建软链接(可以理解为快捷方式)
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh ln - s /usr/local/zookeeper-3.4.8/ /usr/local/zookeeper
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh "ls -l /usr/local/ | grep zookeeper"
这里软链接的用户组合权限可以不用修改，默认为root或者hadoop都可以。

Modify /usr/local/zookeeper/conf/zoo.cfg

I can change a good https://www.lanzous.com/b849762/ Password: 1qq6

批量删除原有的zoo_sample.cfg文件，当然先备份为好
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh rm -f /usr/local/zookeeper/conf/zoo_sample.cfg
把我们准备好的配置文件放进去,批量。
[hadoop@nn1 zk_op]$ ./zk_scp_all.sh ~/zoo.cfg /usr/local/zookeeper/conf/

=================================================================================================

然后修改/usr/local/zookeeper/bin/zkEnv.sh脚本文件，添加日志文件路径
[hadoop@nn1 zk_op]$ vim /usr/local/zookeeper/bin/zkEnv.sh
ZOO_LOG_DIR=/data

把这个配置文件批量分发给其他机器
[hadoop@nn1 zk_op]$ ./zk_scp_all.sh /usr/local/zookeeper/bin/zkEnv.sh /usr/local/zookeeper/bin/

给5台机器创建/data目录，注意这里是给5台机器创建。用的没改过的原本批量脚本。
[hadoop@nn1 hadoop_base_op]$ ./ssh_root.sh mkdir /data
[hadoop@nn1 hadoop_base_op]$ ./ssh_root.sh chown hadoop:hadoop /data
[hadoop@nn1 hadoop_base_op]$ ./ssh_all.sh "ls -l | grep data"

Why is suddenly created above 5 / data behind it ,,, because hdfs and yarn are needed, behind hdfs is running on the three machines, so now are directly created.

然后回到zk_op中，给前三台机器创建id文件。用于zookeeper识别
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh touch /data/myid

Then _ are into three machines _, appended to this document id value.

The first:

echo "1" > /data/myid

Second stage:

echo "2" > /data/myid

Third stage:

echo "3" > /data/myid

3. Batch setting environment variables

在nn1上切换到root用户更改系统环境变量
[hadoop@nn1 zk_op]$ su - root
[root@nn1 ~]# vim /etc/profile
文件在末尾添加
#set Hadoop Path
export HADOOP_HOME=/usr/local/hadoop
export HADOOP_COMMON_HOME=${HADOOP_HOME}
export HADOOP_HDFS_HOME=${HADOOP_HOME}
export HADOOP_MAPRED_HOME=${HADOOP_HOME}
export HADOOP_YARN_HOME=${HADOOP_HOME}
export HADOOP_CONF_DIR=${HADOOP_HOME}/etc/hadoop
export HDFS_CONF_DIR=${HADOOP_HOME}/etc/hadoop
export YARN_CONF_DIR=${HADOOP_HOME}/etc/hadoop
export LD_LIBRARY_PATH=$HADOOP_HOME/lib/native:/usr/lib64

export HBASE_HOME=/usr/local/hbase
export HIVE_HOME=/usr/local/hive
export PATH=$PATH:$JAVA_HOME/bin:$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$HBASE_HOME/bin:$HIVE_HOME/bin:/usr/local/zookeeper/bin

Then the batch is sent to the other two machines

[hadoop@nn1 zk_op]$ ./zk_scp_all.sh /etc/profile /tmp/
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh cp -f /tmp/profile /etc/profile
批量检查一下
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh tail /etc/profile
批量source一下环境变量
[hadoop@nn1 zk_op]$ ./zk_ssh_root.sh source /etc/profile

4. Batch start zookeeper

[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh /usr/local/zookeeper/bin/zkServer.sh start
查看一下是否启动。看看有没有相关进程
[hadoop@nn1 zk_op]$ ./zk_ssh_all.sh jps

View the process as shown below, it shows the process has started successfully QPM

View jps process

Or simply look at the state,

View zookeeper status

Well, zookeeper installation configuration successfully concluded! ! !

-02 to build a Hadoop cluster installation configuration Zookeeper