How to Build a Reliable Zookeeper Cluster on Three Servers

1 Install Zookeeper

Zookeeper can be installed quickly by downloading the binary package, confirming that Java is already installed, and extracting it to the desired location.

Download URL:

https://mirrors.tuna.tsinghua.edu.cn/apache/zookeeper/zookeeper-3.4.8/zookeeper-3.4.8.tar.gz

Copy the archive to each server’s /opt directory and extract:

cd /opt
 tar xzvf zookeeper-3.4.8.tar.gz

2 Configure

Enter the configuration directory, copy the sample file, and edit it:

cd /opt/zookeeper-3.4.8/conf
 cp zoo_sample.cfg zoo.cfg
 vi zoo.cfg

Set the data directory, for example: dataDir=/var/zookeeper Add the three server entries at the end of the file (replace IPs with your own):

server.1=172.17.0.2:2888:3888
server.2=172.17.0.3:2888:3888
server.3=172.17.0.4:2888:3888

Each server.X line defines a unique integer ID for the server.

Create a myid file in the data directory containing the corresponding ID on each host: /var/zookeeper/myid Example contents:

172.17.0.2 → 1
172.17.0.3 → 2
172.17.0.4 → 3

3 Start

Run the start script on all three machines:

/opt/zookeeper-3.4.8/bin/zkServer.sh start

4 Status Check

Verify each node’s status with the status command: /opt/zookeeper-3.4.8/bin/zkServer.sh status Typical successful output is shown in the images below:

5 Operation Test

Execute a few Zookeeper client commands to ensure the ensemble works.

List nodes – the initial state shows only the default /zookeeper node.

Add a node – create a new znode and verify its creation.

Get node data – retrieve the data stored in the newly created node.

Delete the node – remove the znode and confirm deletion.

All tested operations succeed, confirming that the Zookeeper cluster is functioning correctly.