本文演示一主二从。
先说一下遇到的问题。我看网上说想配置哨兵，必须让启动Redis的docker网络模式为host，否则无法访问到从还是什么的。我指定--network host后无法外网访问，redis desktop manager连不上redis了，后来发现我这个新克隆的机器没关防火墙，关上就好了。
1.配置主从
docker pull一下redis
master
mkdir -p /mydata/redis/6379/conf
mkdir -p /mydata/redis/6379/data
touch /mydata/redis/6379/conf/redis.conf
echo "appendonly yes" >> /mydata/redis/6379/conf/redis.conf
docker run --network host --name redis6379 -v /mydata/redis/6379/data:/data -v /mydata/redis/6379/conf/redis.conf:/etc/redis/redis.conf -d redis redis-server /etc/redis/redis.conf --port 6379
slave
#slave1
mkdir -p /mydata/redis/6380/conf
mkdir -p /mydata/redis/6380/data
touch /mydata/redis/6380/conf/redis.conf
echo "appendonly yes" >> /mydata/redis/6380/conf/redis.conf
docker run --network host --name redis6380 -v /mydata/redis/6380/data:/data -v /mydata/redis/6380/conf/redis.conf:/etc/redis/redis.conf -d redis redis-server /etc/redis/redis.conf --port 6380 --slaveof 192.168.243.4 6379
#slave2
mkdir -p /mydata/redis/6381/conf
mkdir -p /mydata/redis/6381/data
touch /mydata/redis/6381/conf/redis.conf
echo "appendonly yes" >> /mydata/redis/6381/conf/redis.conf
docker run --network host --name redis6381 -v /mydata/redis/6381/data:/data -v /mydata/redis/6381/conf/redis.conf:/etc/redis/redis.conf -d redis redis-server /etc/redis/redis.conf --port 6381 --slaveof 192.168.243.4 6379
查看配置：
#主
docker exec -it redis6379 redis-cli
>info replication
#从
docker exec -it redis6380 redis-cli -p 6380
>info replication
2. 配置哨兵
master
mkdir /mydata/redis/6379/sentinel
mkdir /mydata/redis/6379/sentinel/log
vi /mydata/redis/6379/sentinel/sentinel.conf
#输入
port 26379
dir "/var/log/sentinel"
logfile "/var/log/sentinel/26379.log"
sentinel monitor mymaster 192.168.243.4 6379 1 #这里暂时设置成1
#方便看一下主挂掉是什么效果
sentinel down-after-milliseconds mymaster 5000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000
#end
docker run -d --name sentinel26379 -v /mydata/redis/6379/sentinel/sentinel.conf:/conf/sentinel.conf -v /mydata/redis/6379/sentinel/log:/var/log/sentinel --network host redis redis-sentinel /conf/sentinel.conf
先进行测试：
docker exec -it redis6379 redis-cli
> SHUTDOWN
> exit
cat /mydata/redis/6379/sentinel/log/26379.log
此时redis6381已成为新的master
再启动docker start redis6379 6379是slave了
测试成功，修改上面配置文件的sentinel monitor mymaster 192.168.243.4 6379 2 数字配置为2，代表至少有2个Sentinel节点认为主节点不可达，那么这个不可达的判定才是客观的。为了防止票数相同，sentinel启动奇数个。
slave
#slave1
mkdir /mydata/redis/6380/sentinel
mkdir /mydata/redis/6380/sentinel/log
vi /mydata/redis/6380/sentinel/sentinel.conf
#输入
port 26380
dir "/var/log/sentinel"
logfile "/var/log/sentinel/26380.log"
sentinel monitor mymaster 192.168.243.4 6380 2
sentinel down-after-milliseconds mymaster 5000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000
#end
docker run -d --name sentinel26380 -v /mydata/redis/6380/sentinel/sentinel.conf:/conf/sentinel.conf -v /mydata/redis/6380/sentinel/log:/var/log/sentinel --network host redis redis-sentinel /conf/sentinel.conf
#slave2
mkdir /mydata/redis/6381/sentinel
mkdir /mydata/redis/6381/sentinel/log
vi /mydata/redis/6381/sentinel/sentinel.conf
#输入
port 26381
dir "/var/log/sentinel"
logfile "/var/log/sentinel/26381.log"
sentinel monitor mymaster 192.168.243.4 6381 2
sentinel down-after-milliseconds mymaster 5000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000
#end
docker run -d --name sentinel26381 -v /mydata/redis/6381/sentinel/sentinel.conf:/conf/sentinel.conf -v /mydata/redis/6381/sentinel/log:/var/log/sentinel --network host redis redis-sentinel /conf/sentinel.conf
3.SpringBoot连接
application.yml
spring:
redis:
timeout: 5000
sentinel:
master: mymaster
nodes: 192.168.243.4:26379,192.168.243.4:26380,192.168.243.4:26381
controller
@Autowired
private StringRedisTemplate redisTemplate;
@RequestMapping("/redis")
public String redis() {
redisTemplate.opsForValue().set("test", "121323123");
String test = redisTemplate.opsForValue().get("test");
return "RESULT: " + test;
}
访问localhost:8080/redis
到此这篇关于使用Docker配置redis sentinel哨兵的方法步骤的文章就介绍到这了,更多相关Docker配置redis sentinel哨兵内容请搜索我们以前的文章或继续浏览下面的相关文章希望大家以后多多支持我们！
