一、建立三節點 zookeeper 叢集

1. 將 docker-compose.yml 儲存到當前命令列目錄下

docker-compose.yml 檔案

version: '2'
networks:
  zk:
services:
  zookeeper1:
    image: zookeeper
    container_name: zk1.cloud
    networks:
        - zk
    ports:
        - "2181:2181"
    environment:
      ZOO_MY_ID: 1
      ZOO_SERVERS: server.1=0.0.0.0:2888:3888 server.2=zk2.cloud:2888:3888 server.3=zk3.cloud:2888:3888
  zookeeper2:
    image: zookeeper
    container_name: zk2.cloud
    networks:
        - zk
    ports:
        - "2182:2181"
    environment:
      ZOO_MY_ID: 2
      ZOO_SERVERS: server.1=zk1.cloud:2888:3888 server.2=0.0.0.0:2888:3888 server.3=zk3.cloud:2888:3888
  zookeeper3:
    image: zookeeper
    container_name: zk3.cloud
    networks:
        - zk
    ports:
        - "2183:2181"
    environment:
      ZOO_MY_ID: 3
      ZOO_SERVERS: server.1=zk1.cloud:2888:3888 server.2=zk2.cloud:2888:3888 server.3=0.0.0.0:2888:3888
 

2. 執行命令，如果是首次獲取 zookeeper 映象，輸出會有不同

輸入

docker pull zookeeper

輸出

latest: Pulling from library/zookeeper
Digest: sha256:3f03c6f5a91e0f638f3d6a755b2d32c06583766031353be87e2d633fa3006c23
Status: Image is up to date for zookeeper:latest

3. 後臺啟動 zookeeper 叢集

docker-compose up -d

輸出命令

Creating network "dczk_zk" with the default driver
Creating zk2.cloud ... done
Creating zk1.cloud ... done
Creating zk3.cloud ... done

二、檢查程序狀態

輸入

docker ps

輸出

CONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS              PORTS                                        NAMES
cb176c76c620        zookeeper           
 
"/docker-entrypoint.…"   6 minutes ago       Up 6 minutes        2888/tcp, 3888/tcp, 0.0.0.0:2183->2181/tcp   zk3.cloud
ee00aba1c766        zookeeper           "/docker-entrypoint.…"   6 minutes ago       Up 6 minutes        2888/tcp, 0.0.0.0:2181->2181/tcp, 3888/tcp   zk1.cloud
5d95baa693c2        zookeeper           "/docker-entrypoint.…"   6 minutes ago       Up 6 minutes        2888/tcp, 3888/tcp, 0.0.0.0:2182->2181/tcp   zk2.cloud

docker-compose ps

輸出

Name                 Command               State                     Ports
-----------------------------------------------------------------------------------------------
zk1.cloud   /docker-entrypoint.sh zkSe ...   Up      0.0.0.0:2181->2181/tcp, 2888/tcp, 3888/tcp
zk2.cloud   /docker-entrypoint.sh zkSe ...   Up      0.0.0.0:2182->2181/tcp, 2888/tcp, 3888/tcp
zk3.cloud   /docker-entrypoint.sh zkSe ...   Up      0.0.0.0:2183->2181/tcp, 2888/tcp, 3888/tcp

三、測試

1. docker 3 為leader

輸入

docker exec -it zk1.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: follower

輸入

docker exec -it zk2.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: follower

輸入

docker exec -it zk3.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: leader

2. 關閉 zk3.cloud 後檢查leader變化情況

輸入

docker stop zk3.cloud
docker exec -it zk2.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: leader

3. 再次啟動 zk3.cloud 後檢查 leader變化情況

輸入

docker start zk3.cloud
docker exec -it zk2.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: leader

輸入

docker exec -it zk3.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: follower

4. 關閉 zk2.cloud 後檢查 leader變化情況

輸入

docker stop zk2.cloud
docker exec -it zk3.cloud bash ./bin/zkServer.sh status

輸出

ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: leader