怎么用docker搭建redis主从哨兵模式，并整合进springboot项目

这篇文章主要介绍"怎么用docker搭建redis主从哨兵模式，并整合进springboot项目"，在日常操作中，相信很多人在怎么用docker搭建redis主从哨兵模式，并整合进springboot项目问题上存在疑惑，小编查阅了各式资料，整理出简单好用的操作方法，希望对大家解答"怎么用docker搭建redis主从哨兵模式，并整合进springboot项目"的疑惑有所帮助！接下来，请跟着小编一起来学习吧！

一，在主机上搭建redis

整体redis和哨兵架构均为3节点（1主2从）

默认容器启动为bridge模式，这样哨兵取到的主节点ip为容器内网ip，客户端无法访问容器内网，故需要使用宿主机网络模式编排（network_mode:"host"）

宿主机外网IP，192.168.1.254

1，（redis编排文件）master-slave/docker-compose.yml

version: "3"services:  master:    image: redis:latest    container_name: redis-master    command: redis-server --requirepass 123456 --port 6379    ports:    - "6379"    network_mode: "host"  slave1:    image: redis:latest    container_name: redis-slave-1    command: redis-server --port 6380 --slaveof 192.168.1.254 6379 --requirepass 123456 --masterauth 123456    depends_on:    - master    ports:    - "6380"    network_mode: "host"  slave2:    image: redis:latest    container_name: redis-slave-2    command: redis-server --port 6381 --slaveof 192.168.1.254 6379 --requirepass 123456 --masterauth 123456    depends_on:    - master    ports:    - "6381"    network_mode: "host"

在当前目录下使用命令启动：docker-compose up -d

共三台redis，其中master（6379），slave（6380，6381）

2，（哨兵编排文件）sentinel/docker-compose.yml

# Example sentinel.conf can be downloaded from http://download.redis.io/redis-stable/sentinel.confversion: "3"services:  sentinel1:    image: redis:latest    container_name: redis-sentinel-1    command: redis-sentinel /usr/local/etc/redis/sentinel.conf    ports:    - "26379"    volumes:    - "/root/redis/sentinel1.conf:/usr/local/etc/redis/sentinel.conf"    network_mode: "host"  sentinel2:    image: redis:latest    container_name: redis-sentinel-2    command: redis-sentinel /usr/local/etc/redis/sentinel.conf    ports:    - "26380"    volumes:    - "/root/redis/sentinel2.conf:/usr/local/etc/redis/sentinel.conf"    network_mode: "host"  sentinel3:    image: redis:latest    container_name: redis-sentinel-3    command: redis-sentinel /usr/local/etc/redis/sentinel.conf    ports:    - "26381"    volumes:    - "/root/redis/sentinel3.conf:/usr/local/etc/redis/sentinel.conf"    network_mode: "host"

在当前目录下使用命令启动：docker-compose up -d

共三台哨兵，端口分别为26379， 26380， 26381

其中三个哨兵配置文件内容除了端口不同，其他内容均相同

/root/redis/sentinel1.conf

port 26379 # 其他两节点分别为26380, 26381dir /tmpsentinel monitor mymaster 192.168.1.254 6379 2sentinel auth-pass mymaster 123456sentinel down-after-milliseconds mymaster 30000sentinel parallel-syncs mymaster 1sentinel failover-timeout mymaster 10000sentinel deny-scripts-reconfig yes

#第三行表示Redis监控一个叫做mymaster的运行在192.168.1.254:6379的master，投票达到2则表示master以及挂掉了。
#第四行设置主节点的认证密码
#第五行表示在一段时间范围内sentinel向master发送的心跳PING没有回复则认为master不可用了。
#第六行的parallel-syncs表示设置在故障转移之后，同时可以重新配置使用新master的slave的数量。数字越低，更多的时间将会用故障转移完成，但是如果slaves配置为服务旧数据，你可能不希望所有的slave同时重新同步master。因为主从复制对于slave是非阻塞的，当停止从master加载批量数据时有一个片刻延迟。通过设置选项为1，确信每次只有一个slave是不可到达的。
#第七行表示10秒内mymaster还没活过来，则认为master宕机了。

这是启动后的容器列表：

二，整合springboot项目

1，引入依赖

                                   org.springframework.boot                        spring-boot-starter-data-redis                

2，配置application.yml

spring:  redis:    sentinel:      master: mymaster      nodes:      - "192.168.1.254:26379"      - "192.168.1.254:26380"      - "192.168.1.254:26381"    host: 192.168.1.254    password: 123456    jedis:      pool:        min-idle: 8        max-active: 100        max-wait: 3000        max-idle: 100

3，使用RedisTemplate操作redis

    @Autowired        RedisTemplate redisTemplate;                @Test        public void testRedisMasterSlave() throws Exception {                ValueOperations valueOperations = redisTemplate.opsForValue();                ExecutorService es = Executors.newFixedThreadPool(3);                for (int j = 0; j < 3; j++) {                        es.submit(() -> {                                for (int i = 0; i < 1000; i++) {                                        try {                                                String threadName = Thread.currentThread().getName();                                                valueOperations.set(threadName+i, i+"", 30L, TimeUnit.MINUTES);                                                TimeUnit.MILLISECONDS.sleep(200L);                                        } catch (InterruptedException e) {                                                System.out.println("error: " + e.getMessage());                                        }                                }                        });                }                es.shutdown();                es.awaitTermination(30L, TimeUnit.MINUTES);        }

这里使用3个线程模拟并发写入redis，在写的过程中，停掉master，查看控制台输出日志

2019-10-06 18:52:51.949  INFO 128972 --- [pool-1-thread-1] io.lettuce.core.EpollProvider            : Starting without optional epoll library2019-10-06 18:52:51.951  INFO 128972 --- [pool-1-thread-1] io.lettuce.core.KqueueProvider           : Starting without optional kqueue library2019-10-06 18:53:55.979  INFO 128972 --- [xecutorLoop-1-8] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was /192.168.1.254:63792019-10-06 18:53:57.992  WARN 128972 --- [ioEventLoop-4-4] i.l.core.protocol.ConnectionWatchdog     : Cannot reconnect: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: no further information: /192.168.1.254:63792019-10-06 18:54:02.277  INFO 128972 --- [xecutorLoop-1-7] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was 192.168.1.254:63792019-10-06 18:54:04.286  WARN 128972 --- [ioEventLoop-4-2] i.l.core.protocol.ConnectionWatchdog     : Cannot reconnect: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: no further information: /192.168.1.254:63792019-10-06 18:54:08.577  INFO 128972 --- [xecutorLoop-1-4] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was 192.168.1.254:63792019-10-06 18:54:10.587  WARN 128972 --- [ioEventLoop-4-8] i.l.core.protocol.ConnectionWatchdog     : Cannot reconnect: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: no further information: /192.168.1.254:63792019-10-06 18:54:15.678  INFO 128972 --- [xecutorLoop-1-3] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was 192.168.1.254:63792019-10-06 18:54:17.687  WARN 128972 --- [ioEventLoop-4-6] i.l.core.protocol.ConnectionWatchdog     : Cannot reconnect: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: no further information: /192.168.1.254:63792019-10-06 18:54:22.877  INFO 128972 --- [xecutorLoop-1-7] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was 192.168.1.254:63792019-10-06 18:54:24.885  WARN 128972 --- [ioEventLoop-4-2] i.l.core.protocol.ConnectionWatchdog     : Cannot reconnect: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: no further information: /192.168.1.254:63792019-10-06 18:54:29.078  INFO 128972 --- [xecutorLoop-1-1] i.l.core.protocol.ConnectionWatchdog     : Reconnecting, last destination was 192.168.1.254:63792019-10-06 18:54:29.085  INFO 128972 --- [ioEventLoop-4-4] i.l.core.protocol.ReconnectionHandler    : Reconnected to 192.168.1.254:6380

可以看到在18:53:55.979，已经连不上主节点了（6379），之后一直在重连，直到18:54:29.085才连上，但是这时候的节点已经变了（6380），证明哨兵已经可以自动切换主节点了。

再看redis中的数据，一共3000条数据，一条没丢，证明即使在切换主备的过程中，数据也不会丢

接着启动刚停掉的redis-master节点，发现已经启动不了了，报错

2019-10-05T16:00:57.899358510Z 1:S 05 Oct 2019 16:00:57.897 * Connecting to MASTER 192.168.1.254:63802019-10-05T16:00:57.899381586Z 1:S 05 Oct 2019 16:00:57.897 * MASTER <-> REPLICA sync started2019-10-05T16:00:57.899385377Z 1:S 05 Oct 2019 16:00:57.897 * Non blocking connect for SYNC fired the event.2019-10-05T16:00:57.899388194Z 1:S 05 Oct 2019 16:00:57.898 * Master replied to PING, replication can continue...2019-10-05T16:00:57.899391119Z 1:S 05 Oct 2019 16:00:57.898 * (Non critical) Master does not understand REPLCONF listening-port: -NOAUTH Authentication required.2019-10-05T16:00:57.899401303Z 1:S 05 Oct 2019 16:00:57.898 * (Non critical) Master does not understand REPLCONF capa: -NOAUTH Authentication required.2019-10-05T16:00:57.899404228Z 1:S 05 Oct 2019 16:00:57.898 * Partial resynchronization not possible (no cached master)2019-10-05T16:00:57.899406828Z 1:S 05 Oct 2019 16:00:57.898 # Unexpected reply to PSYNC from master: -NOAUTH Authentication required.2019-10-05T16:00:57.899409536Z 1:S 05 Oct 2019 16:00:57.898 * Retrying with SYNC...2019-10-05T16:00:57.899412136Z 1:S 05 Oct 2019 16:00:57.899 # MASTER aborted replication with an error: NOAUTH Authentication required.

这时候已经作为从节点连接新主节点了，但是授权认证失败，因为之前主节点根本就没有配置slaveof，这里我以为哨兵会自动加上slaveof配置，但是没有，难道是没有使用配置文件的缘故？

已经找到原因，并且在这篇中完全实现了

到此，关于"怎么用docker搭建redis主从哨兵模式，并整合进springboot项目"的学习就结束了，希望能够解决大家的疑惑。理论与实践的搭配能更好的帮助大家学习，快去试试吧！若想继续学习更多相关知识，请继续关注网站，小编会继续努力为大家带来更多实用的文章！