使用docker-compose搭建zookeeper集群环境

使用docker-compose搭建zookeeper集群环境

zookeeper是一个集群环境，用来管理微服务架构下面的配置管理功能。

这篇文章是一个基础步骤如何搭建zookeeper的docker集群环境。
我们使用docker-compose来搭建如下的zookeeper集群环境：

1. 集群包含三个node：zookeeper1, zookeeper2, zookeeper3

1. 下载zookeeper docker image

$ docker pull zookeeper

2. 编辑docker-compose.yaml文件

version: '2'
networks:
  byfn:

services:
  zookeeper1:
    image: zookeeper
    container_name: zookeeper1
    environment:
        - ZOO_MY_ID=1
        - ZOO_SERVERS=server.1=zookeeper1:2888:3888 server.2=zookeeper2:2888:3888 server.3=zookeeper3:2888:3888
    ports:
      - 2181
    networks:
      - byfn

  zookeeper2:
    image: zookeeper
    container_name: zookeeper2
    environment:
        - ZOO_MY_ID=2
        - ZOO_SERVERS=server.1=zookeeper1:2888:3888 server.2=zookeeper2:2888:3888 server.3=zookeeper3:2888:3888
    ports:
      - 2181
    networks:
      - byfn

  zookeeper3:
    image: zookeeper
    container_name: zookeeper3
    environment:
        - ZOO_MY_ID=3
        - ZOO_SERVERS=server.1=zookeeper1:2888:3888 server.2=zookeeper2:2888:3888 server.3=zookeeper3:2888:3888
    ports:
      - 2181
    networks:
      - byfn

详细的配置请参考文档：https://hub.docker.com/_/zookeeper/
这里非常清楚的描述了如何修改配置，存储数据，以及调整日志级别以及输出等设置。

3. 启动服务

$ docker-compose up

4. 验证集群的状态

$ docker ps
CONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS              PORTS                                         NAMES
c1f7646bf0e6        zookeeper           "/docker-entrypoin..."   23 minutes ago      Up 23 minutes       2888/tcp, 3888/tcp, 0.0.0.0:32794->2181/tcp   zookeeper2
7e05f06c258d        zookeeper           "/docker-entrypoin..."   23 minutes ago      Up 23 minutes       2888/tcp, 3888/tcp, 0.0.0.0:32793->2181/tcp   zookeeper3
d293438c1813        zookeeper           "/docker-entrypoin..."   23 minutes ago      Up 23 minutes       2888/tcp, 3888/tcp, 0.0.0.0:32792->2181/tcp   zookeeper1

$ docker exec -t zookeeper1 zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: follower
$ docker exec -t zookeeper2 zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: follower
$ docker exec -t zookeeper3 zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /conf/zoo.cfg
Mode: leader

这里我们看到三个zookeeper节点，zookeeper3是leader，zookeeper1和zookeeper2是follower。

其他状态检查命令：

$ nc localhost 32794 <<< ruok
imok

$ echo srvr | nc localhost 32794
Zookeeper version: 3.4.13-2d71af4dbe22557fda74f9a9b4309b15a7487f03, built on 06/29/2018 04:05 GMT
Latency min/avg/max: 0/7/116
Received: 40
Sent: 39
Connections: 1
Outstanding: 0
Zxid: 0x10000000e
Mode: follower
Node count: 5

4. 集群的使用

4.1 基本操作

[zk: localhost:2181(CONNECTED) 4] create /foo "Hello foo"
Created /foo

[zk: localhost:2181(CONNECTED) 5] get /foo
Hello foo
cZxid = 0x100000013
ctime = Fri Jan 04 02:29:47 GMT 2019
mZxid = 0x100000013
mtime = Fri Jan 04 02:29:47 GMT 2019
pZxid = 0x100000013
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 9
numChildren = 0

[zk: localhost:2181(CONNECTED) 6] ls /foo 
[]

4.2 创建路径create

[zk: localhost:2181(CONNECTED) 7] create /foo1/foo1 "Hello foo1"
Node does not exist: /foo1/foo1

创建失败，因为zookeeper不支持递归创建，必须先一级一级的创建父目录节点。
创建父目录时，让值为空就可以。(其实目录也可以带值，两者并不冲突。)

[zk: localhost:2181(CONNECTED) 14] create /foo1 ""     # 或者：create /foo1 ""
Created /foo1

[zk: localhost:2181(CONNECTED) 15] create /foo1/foo1 "Hello foo1"
Created /foo1/foo1

4.3 查看节点的元素ls

[zk: localhost:2181(CONNECTED) 21] ls /foo1     
[foo1]

因为/foo1是一个目录，他下面有一个子节点/foo1/foo1

[zk: localhost:2181(CONNECTED) 21] ls /foo1/foo1     
[]

因为/foo1/foo1是一个元节点，他下面没有子节点了。

注意，ls命令只能列出一级目录，不能递归列出所有的子目录。

4.4 获取值get

get /foo1/foo1

4.5 修改值set

set /foo1/foo1 'newvalue'

4.6 删除节点delete

delete /foo1/foo1

注意只能删除子节点为空的节点；所以如果要删除一个节点，必须先把所有的子节点都删除。
如果要删除非空目录，还可以使用rmr命令。