How to build a Zookeeper cluster in CentOS

Deploying a ZooKeeper cluster on a CentOS system requires the following steps:

Environmental preparation

1. Install the Java Runtime Environment: Use the following command to install the Java 8 development kit:

   sudo yum install java-1.8.0-openjdk-devel

2. Download ZooKeeper: Download the version for CentOS (such as ZooKeeper 3.8.x) from the official Apache ZooKeeper website. Use the wget command to download and replace zookeeper-3.8.x with the actual version number:

   wget https://downloads.apache.org/zookeeper/zookeeper-3.8.x/zookeeper-3.8.x.tar.gz

3. Unzip ZooKeeper: Unzip the downloaded compressed package and enter the decompressed directory:

   tar -zxvf zookeeper-3.8.x.tar.gz
   cd zookeeper-3.8.x

Configure ZooKeeper

1. Create a configuration file: Copy the sample configuration file and rename it to zoo.cfg :

   cp conf/zoo_sample.cfg conf/zoo.cfg

2. Edit the configuration file ( zoo.cfg ): Use a text editor (such as vi or nano ) to edit the zoo.cfg file and configure the following parameters:

   • tickTime : Basic time unit (milliseconds), recommended value 2000.
   • dataDir : ZooKeeper data storage directory, such as /var/lib/zookeeper . Make sure that the directory exists and that the ZooKeeper process has read and write permissions.
   • clientPort : Client connection port, default 2181.
   • initLimit : The maximum tolerant number of heartbeat intervals when initializing the connection, the recommended value is 10.
   • syncLimit : The length of time to send messages, requests and responses between the Leader and the Follower, the recommended value is 5.
   • server.x : Assign a unique ID and address to each server in the cluster. For example, the three-node cluster configuration is as follows:

     <code>server.1=192.168.1.1:2888:3888 server.2=192.168.1.2:2888:3888 server.3=192.168.1.3:2888:3888</code>

     Please replace the IP address with the actual server IP address. The myid file of each server (in the dataDir directory) must contain its corresponding server ID (1, 2, 3).

Start ZooKeeper cluster

1. Create myid file: Create a file named myid in the dataDir directory of each server and write the server ID (1, 2, 3) to the file. For example, on server 1, the content of myid file is 1 .

2. Start the ZooKeeper service: Execute the following commands on each server to start the ZooKeeper service, and use the server ID as the parameter:

   ./bin/zkServer.sh start

3. Verify cluster status: Use the following command to check cluster status:

   ./bin/zkServer.sh status

   If the cluster starts successfully, you should see that all servers are running. You can use ./bin/zkCli.sh to connect to the ZooKeeper cluster and execute commands to test it.

Set up the power-on self-start (optional)

For easy management, it is recommended to set ZooKeeper to start up automatically. This part of the steps is relatively complicated. It is recommended to refer to the official ZooKeeper document or use the service management tools provided by the system for configuration.

Important notes

• Time synchronization: Ensure time synchronization of all servers, you can use NTP services.
• Firewall: Ensure that the firewall allows network communications to the ports (2181, 2888, 3888) used by ZooKeeper.
• myid file: The server.x configuration in the myid file and the zoo.cfg file must be consistent and correct.
• Directory permissions: Make sure the ZooKeeper data directory has the correct permissions.

Through the above steps, you can successfully build a ZooKeeper cluster on CentOS. Please note that the IP address and version number of the placeholder in the text are your actual value.

The above is the detailed content of How to build a Zookeeper cluster in CentOS. For more information, please follow other related articles on the PHP Chinese website!