Skip to content

Latest commit

 

History

History
280 lines (199 loc) · 6.34 KB

docker_compose.md

File metadata and controls

280 lines (199 loc) · 6.34 KB
Raw

ZooKeeper Cluster(on a single node) Setup with Docker Compose

Environment:

  • Red Hat Enterprise Linux Server release 6.6 (Santiago)

  • Kernel: 2.6.32-504.23.4.el6.x86_64

  • Docker: Server version 1.7.1 Storage Driver devicemapper Execution Driver native-0.2

  • Docker Compose:

  • ZooKeeper: 3.4.6-1569965, built on 02/20/2014 09:09 GMT

Docker Compose is an orchestration tool that makes spinning up multi-container applications effortless. Dockerfile makes user define and manage a single container, while Compose helps user to define a group of containers with dependency in a single yaml file, and

0. Install docker

Follow the installation guide to install docker in your system

1. Install Docker compose 1.4.2

1.1 Enter the curl command in your termial.

The command has the following format:

curl -L https://github.com/docker/compose/releases/download/1.4.2/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

1.2 Apply executable permissions to the binary:

chmod +x /usr/local/bin/docker-compose

1.3 Test the installation

~ docker-compose --version
docker-compose version: 1.4.2
2. Define the ZooKeeper cluster in docker compose

2.1 Create common.yml to define the common sections of ZooKeeper cluster members

zookeeper:
  image: garland/zookeeper
  net: "host"
  environment:
    - ADDITIONAL_ZOOKEEPER_1=server.1=localhost:2888:3888
    - ADDITIONAL_ZOOKEEPER_2=server.2=localhost:2889:3889
    - ADDITIONAL_ZOOKEEPER_3=server.3=localhost:2890:3890

2.2 Create docker-compose.yml to define the cluster

zk1:
  extends:
    file: common.yml
    service: zookeeper
  volumes:
    - /data/zk1:/tmp/zookeeper
  environment:
    - SERVER_ID=1
    - ADDITIONAL_ZOOKEEPER_4=clientPort=2181
zk2:
  extends:
    file: common.yml
    service: zookeeper
  volumes:
    - /data/zk2:/tmp/zookeeper
  environment:
    - SERVER_ID=2
    - ADDITIONAL_ZOOKEEPER_4=clientPort=2182
zk3:
  extends:
    file: common.yml
    service: zookeeper
  volumes:
    - /data/zk2:/tmp/zookeeper
  environment:
    - SERVER_ID=3
    - ADDITIONAL_ZOOKEEPER_4=clientPort=2183
3. Start ZooKeeper Cluster with docker compose

3.1 Put common.yml and docker-compose.yml into a directory, for example

~ pwd
/zookeeper_compose
~ ls
common.yml  docker-compose.yml

3.2 Start the containers with docker compose

~ docker-compose up -d
Creating zookeepercompose_zk1_1...
Creating zookeepercompose_zk2_1...
Creating zookeepercompose_zk3_1...

3.3 Check the status

~ docker-compose ps
         Name              Command     State   Ports
----------------------------------------------------
zookeepercompose_zk1_1   /opt/run.sh   Up
zookeepercompose_zk2_1   /opt/run.sh   Up
zookeepercompose_zk3_1   /opt/run.sh   Up
4. Test ZooKeeper Cluster

4.1 Create a new znode in cluster member zookeepercompose_zk1_1

~ docker exec -it zookeepercompose_zk1_1 bash
~ /opt/zookeeper/bin/zkCli.sh -server 127.0.0.1:2181 (inside container zookeepercompose_zk1_1)
[zk: 127.0.0.1:2181(CONNECTED) 0]create /test1 abc
Connecting to 127.0.0.1:2181
Created /test1

4.2 Check the znode in cluster member zookeepercompose_zk2_1

~ docker exec -it zookeepercompose_zk2_1 bash
~ /opt/zookeeper/bin/zkCli.sh -server 127.0.0.1:2182 (inside container zookeepercompose_zk2_1)
Connecting to 127.0.0.1:2182
[zk: 127.0.0.1:2182(CONNECTED) 0] get /test1
abc
[zk: 127.0.0.1:2182(CONNECTED) 0] get /test1
abc
cZxid = 0x100000004
ctime = Tue Oct 13 14:44:04 UTC 2015
mZxid = 0x100000004
mtime = Tue Oct 13 14:44:04 UTC 2015
pZxid = 0x100000004
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 3
numChildren = 0
5. Develop java client ZooKeeper Cluster

5.1 Develop the code

package org.cheyang.zookeeper.demo;

import java.io.IOException;

import org.apache.zookeeper.CreateMode;
import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooDefs.Ids;
import org.apache.zookeeper.ZooKeeper;

public class Demo {
	
	public static void main(String[] args) throws  IOException, KeeperException, InterruptedException{
		
	
		ZooKeeper zk = new ZooKeeper("localhost:2181,localhost:2182,localhost:2183", 60000, new Watcher() {
            // monitor all the events
            public void process(WatchedEvent event) {
                System.out.println("EVENT:" + event.getType());
            }
        });
		
		String key = "/test";
		
		String value = "abc";
		
		
		// create or get znode
		if(zk.exists(key, false) == null){
			
			String createPath = zk.create(key, value.getBytes(),  
                    Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);  

		    System.out.println("Created znode:"+createPath);  
			
		}else{
			byte[] data = zk.getData(key, false, null);
			
			System.out.print("Get znode:"); 
			System.out.printf("[%s,%s]", key, new String(data)); 
			System.out.println(""); 
		}

        // check root znode
        System.out.println("ls / => " + zk.getChildren("/", true));

        zk.close();
	}
	
	
}

5.2 compile the code

~  javac -cp .:/opt/zookeeper/zookeeper-3.4.6.jar org/zookeeper/demo/Demo.java

5.3 execute the code

~ java -cp /opt/zookeeper/zookeeper-3.4.6.jar:/opt/zookeeper/lib/*:.  org.zookeeper.demo.Demo

5.4 stop one cluster member

~ /opt/zookeeper/bin/zkServer.sh stop zk1.cfg

5.5 execute the code again

~ java -cp /opt/zookeeper/zookeeper-3.4.6.jar:/opt/zookeeper/lib/*:.  org.zookeeper.demo.Demo
6. Delete ZooKeeper Cluster when finished using it
~  docker-compose stop
Stopping zookeepercompose_zk3_1... done
Stopping zookeepercompose_zk2_1... done
Stopping zookeepercompose_zk1_1... done

~  docker-compose rm -f
Going to remove zookeepercompose_zk3_1, zookeepercompose_zk2_1, zookeepercompose_zk1_1
Removing zookeepercompose_zk3_1... done
Removing zookeepercompose_zk2_1... done
Removing zookeepercompose_zk1_1... done

Pros

  1. easy to automate deployment, to start, to stop, to destroy
  2. portable

Cons

  1. Have the same weakness as docker
  2. Docker team doesn't suggest to put it in production