Docker 之部署Redis集群(二)

前言

本小节我们学习一下使用Dokcer部署Redis集群之后为确保Redis HA性能实现Sentinel模式。
如何使用Docker部署Redis集群。

案例

参照前文，假设你已经成功部署了Redis集群。

• 目录参考

  [root@iZ2ze8sm5upgi8z1dcazqeZ redis]# tree
  .
  ├── data
  │   ├── appendonly.aof
  │   ├── docker-compose.yml
  │   └── dump.rdb
  └── sentinel
      ├── config
      │   ├── docker-compose.yml
      │   ├── sentinel1.conf
      │   ├── sentinel2.conf
      │   ├── sentinel3.conf
      │   └── sentinel.conf
      └── data
  
  4 directories, 8 files
  [root@iZ2ze8sm5upgi8z1dcazqeZ redis]# 
  

• 在中workspace/redis/sentinel/config编写sentinel.conf文件

  port 26379
  dir /workspace/redis/sentinel/config
  # 自定义集群名，其中 182.92.0.228 为 redis-master 的 ip，6379 为 redis-master 的端口，2 为最小投票数（因为有 3 台 Sentinel 所以可以设置成 2）
  sentinel monitor redis-master 182.92.0.228 6379 2
  sentinel down-after-milliseconds redis-master 30000
  sentinel parallel-syncs redis-master 1
  sentinel auth-pass redis-master Dustyone
  sentinel failover-timeout redis-master 180000
  sentinel deny-scripts-reconfig yes
  

• 将sentinel.conf复制三份，复制文件与sentinel.conf同路径

  cp sentinel.conf sentinel1.conf
  cp sentinel.conf sentinel2.conf
  cp sentinel.conf sentinel3.conf
  

• 编写docker-composr.yml

  version: '3.7'
  services:
    sentinel1:
      image: redis
      container_name: redis-sentinel-1
      command: redis-sentinel /workspace/redis/sentinel/config/sentinel.conf
      restart: always
      ports:
        - 26379:26379
      volumes:
        - /workspace/redis/sentinel/config/sentinel1.conf:/workspace/redis/sentinel/config/sentinel.conf
  
    sentinel2:
      image: redis
      container_name: redis-sentinel-2
      command: redis-sentinel /workspace/redis/sentinel/config/sentinel.conf
      restart: always
      ports:
        - 26380:26380
      volumes:
        - /workspace/redis/sentinel/config/sentinel2.conf:/workspace/redis/sentinel/config/sentinel.conf
  
    sentinel3:
      image: redis
      container_name: redis-sentinel-3
      command: redis-sentinel /workspace/redis/sentinel/config/sentinel.conf
      restart: always
      ports:
        - 26381:26381
      volumes:
        - /workspace/redis/sentinel/config/sentinel3.conf:/workspace/redis/sentinel/config/sentinel.conf
  

• 在/workspace/redis/sentinel/config启动docker-compose.yml

  docker-compose up -d
  

• 查看redis-sentinel是否启动成功

  [root@iZ2ze8sm5upgi8z1dcazqeZ config]# docker ps -a
  CONTAINER ID        IMAGE                             COMMAND                  CREATED             STATUS              PORTS                                            NAMES
  72eb47246eec        redis                             "docker-entrypoint.s…"   6 minutes ago       Up 6 minutes        6379/tcp, 0.0.0.0:26380->26380/tcp               redis-sentinel-2
  c68c5d3c1ecf        redis                             "docker-entrypoint.s…"   6 minutes ago       Up 6 minutes        6379/tcp, 0.0.0.0:26381->26381/tcp               redis-sentinel-3
  a5df697f45bd        redis                             "docker-entrypoint.s…"   6 minutes ago       Up 6 minutes        6379/tcp, 0.0.0.0:26379->26379/tcp               redis-sentinel-1
  a43bcfd77e5a        redis                             "docker-entrypoint.s…"   2 hours ago         Up 2 hours          6379/tcp, 0.0.0.0:6380->6380/tcp                 redis-slave-1
  19b10ac59405        redis                             "docker-entrypoint.s…"   2 hours ago         Up 2 hours          6379/tcp, 0.0.0.0:6381->6381/tcp                 redis-slave-2
  9acdc4bd4086        redis                             "docker-entrypoint.s…"   2 hours ago         Up 16 minutes       0.0.0.0:6379->6379/tcp                           redis-master
  

• 进入redis-sentinel容器中,查看主redis连接状态

  [root@iZ2ze8sm5upgi8z1dcazqeZ config]# docker exec -it redis-sentinel-1 bash
  root@a5df697f45bd:/data# redis-cli -p 26379
  127.0.0.1:26379> sentinel master redis-master
   1) "name"
   2) "redis-master"
   3) "ip"
   4) "182.92.0.228"
   5) "port"
   6) "6380"
   7) "runid"
   8) "4ab8f3fb673651053d3750dd8a72d98f1bc872c1"
   9) "flags"
  10) "master"
  11) "link-pending-commands"
  12) "0"
  13) "link-refcount"
  14) "1"
  15) "last-ping-sent"
  16) "0"
  17) "last-ok-ping-reply"
  18) "858"
  19) "last-ping-reply"
  20) "858"
  21) "down-after-milliseconds"
  22) "30000"
  23) "info-refresh"
  24) "8224"
  25) "role-reported"
  26) "master"
  27) "role-reported-time"
  28) "520683"
  29) "config-epoch"
  30) "1"
  31) "num-slaves"
  32) "2"
  33) "num-other-sentinels"
  34) "2"
  35) "quorum"
  36) "2"
  37) "failover-timeout"
  38) "180000"
  39) "parallel-syncs"
  40) "1"
  127.0.0.1:26379> 
  

• 查看从redis信息是否正常

  127.0.0.1:26379> sentinel slaves redis-master
  1)  1) "name"
      2) "182.92.0.228:6381"
      3) "ip"
      4) "182.92.0.228"
      5) "port"
      6) "6381"
      7) "runid"
      8) "54c6883557cc0cbb28e521997bb1b25ddd2908d9"
      9) "flags"
     10) "slave"
     11) "link-pending-commands"
     12) "0"
     13) "link-refcount"
     14) "1"
     15) "last-ping-sent"
     16) "0"
     17) "last-ok-ping-reply"
     18) "789"
     19) "last-ping-reply"
     20) "789"
     21) "down-after-milliseconds"
     22) "30000"
     23) "info-refresh"
     24) "1724"
     25) "role-reported"
     26) "slave"
     27) "role-reported-time"
     28) "644775"
     29) "master-link-down-time"
     30) "0"
     31) "master-link-status"
     32) "ok"
     33) "master-host"
     34) "182.92.0.228"
     35) "master-port"
     36) "6380"
     37) "slave-priority"
     38) "100"
     39) "slave-repl-offset"
     40) "432646"
  2)  1) "name"
      2) "182.92.0.228:6379"
      3) "ip"
      4) "182.92.0.228"
      5) "port"
      6) "6379"
      7) "runid"
      8) "e286d7fa7661b9c9cf4c984ee36ddb69a21ccc5b"
      9) "flags"
     10) "slave"
     11) "link-pending-commands"
     12) "0"
     13) "link-refcount"
     14) "1"
     15) "last-ping-sent"
     16) "0"
     17) "last-ok-ping-reply"
     18) "623"
     19) "last-ping-reply"
     20) "623"
     21) "down-after-milliseconds"
     22) "30000"
     23) "info-refresh"
     24) "623"
     25) "role-reported"
     26) "slave"
     27) "role-reported-time"
     28) "644775"
     29) "master-link-down-time"
     30) "1569469175000"
     31) "master-link-status"
     32) "err"
     33) "master-host"
     34) "182.92.0.228"
     35) "master-port"
     36) "6380"
     37) "slave-priority"
     38) "100"
     39) "slave-repl-offset"
     40) "1"
  127.0.0.1:26379> 
  

• 测试主redis挂了之后,哨兵能否正常选举redis

  • 停掉主redis

    docker stop redis-master
    

  • 查看redis-sentinel日志,看其将那个redis选举为主

    [root@iZ2ze8sm5upgi8z1dcazqeZ config]# docker logs redis-sentinel-2
    1:X 26 Sep 2019 03:28:51.588 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
    1:X 26 Sep 2019 03:28:51.588 # Redis version=5.0.6, bits=64, commit=00000000, modified=0, pid=1, just started
    1:X 26 Sep 2019 03:28:51.588 # Configuration loaded
    1:X 26 Sep 2019 03:28:51.590 * Running mode=sentinel, port=26379.
    1:X 26 Sep 2019 03:28:51.590 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.
    1:X 26 Sep 2019 03:28:51.590 # Sentinel ID is 02dd0257960310b300feef92f56d484f55df4b44
    1:X 26 Sep 2019 03:28:51.590 # +monitor master redis-master 182.92.0.228 6380 quorum 2
    1:X 26 Sep 2019 03:28:53.707 * +sentinel-address-switch master redis-master 182.92.0.228 6380 ip 172.26.0.4 port 26379 for e41d3841f9c208262b8564d1678050b8e8fb2e06
    1:X 26 Sep 2019 03:28:53.738 * +sentinel-address-switch master redis-master 182.92.0.228 6380 ip 172.26.0.3 port 26379 for 8be5d0ebdebe9b14139ac7db9b4d9d97f613e3e2
    1:X 26 Sep 2019 03:43:50.239 # +sdown slave 182.92.0.228:6379 182.92.0.228 6379 @ redis-master 182.92.0.228 6380
    [root@iZ2ze8sm5upgi8z1dcazqeZ config]# 
    

    如上可知，6380上的Redis被选举为主Redis了。

  • 测试端口号为6380的redis是否具备写功能,如下则表明成功

    [root@iZ2ze8sm5upgi8z1dcazqeZ config]# docker exec -it redis-slave-1 bash
    root@a43bcfd77e5a:/data# redis-cli -p 6380
    127.0.0.1:6380> auth Dustyone
    OK
    127.0.0.1:6380> keys *
    1) "reob"
    2) "broker"
    3) "sales"
    127.0.0.1:6380> set buyers Duyb
    OK
    127.0.0.1:6380> keys *
    1) "reob"
    2) "broker"
    3) "buyers"
    4) "sales"
    127.0.0.1:6380> get buyers
    "Duyb"
    127.0.0.1:6380> 
    

如上便检验成功了

小节

• 本案例实现为单机版集群，如若主从不在同一服务器上，请设置合理的主从之间的心跳检测时间
• 场景：本案例中将6379的Redis设置成主Redis，当主Redis挂了之后在重新加入集群时，不再是Master角色了。
• 开发的安全组