Redis哨兵模式（Sentinel）的搭建与配置-编程学习网

Redis 哨兵模式（Sentinel）是一个自动监控处理 redis 间故障节点转移工作的一个redis服务端实例，它不提供数据存储服务，只进行普通 redis 节点监控管理，使用redis哨兵模式可以实现redis服务端故障的自动化转移。

一、搭建redis主从集群

1、创建3个redis实例

关于redis的搭建，可以参考历史文章。

https://mp.weixin.qq.com/s/RaWy0sqRxcAti1qbv-GbZQ

如果有编译好的二进制文件，则直接部署redis实例即可。

创建三个redis实例所需的目录，生产环境需独立部署在不同主机上，提高稳定性。

mkdir   -p /data/redis
cd /data/redis/
mkdir  redis6379  redis6380  redis6381
cd redis6379
vim redis.conf
#  添加如下配置
bind  0.0.0.0
protected-mode no
port 6379
tcp-backlog 511
timeout 30
tcp-keepalive 300
daemonize yes
supervised no
pidfile /data/redis/redis6379/redis_6379.pid
loglevel notice
logfile "/data/redis/redis6379/redis6379.log"
databases 16
stop-writes-on-bgsave-error yes
rdbcompression yes
rdbchecksum yes
dbfilename dump.rdb
dir /data/redis/redis6379
masterauth 123456
slave-serve-stale-data yes
slave-read-only yes
# 将配置文件拷贝到其他2个实例的目录下
cp redis.conf  ../redis6380.conf
cp redis.conf ../redis6381.conf
sed -i  "s#6379#6380#g"  ../redis6380/redis.conf
sed -i  "s#6379#6381#g"  ../redis6381/redis.conf
# redis实例不建议使用root账号启动，单独创建一个redis用户，并修改redis相关目录的权限
useradd redis
chown -R   redis:redis /data/redis
 su - redis
#  启动三个redis实例
redis-server  /data/redis/redis6379/redis.conf
redis-server  /data/redis/redis6380/redis.conf
redis-server  /data/redis/redis6381/redis.conf

2、配置主从同步

进入2个实例，配置同步，配置完成后去主节点检查一下是否正常。

[redis@test redis6379]$ redis-cli -p 6380 -a "123456"
Warning: Using a password with '-a' option on the command line interface may not be safe.
127.0.0.1:6380> slaveof 127.0.0.1 6379
OK
127.0.0.1:6380> exit
[redis@test redis6379]$ redis-cli -p 6381 -a "123456"
Warning: Using a password with '-a' option on the command line interface may not be safe.
127.0.0.1:6381> slaveof 127.0.0.1 6379
OK
127.0.0.1:6381> exit
[redis@test redis6379]$ redis-cli -p 6379 -a "123456"
Warning: Using a password with '-a' option on the command line interface may not be safe.
127.0.0.1:6379> info replication
# Replication
role:master
connected_slaves:2
slave0:ip=127.0.0.1,port=6380,state=online,offset=42,lag=0
slave1:ip=127.0.0.1,port=6381,state=online,offset=42,lag=1
master_replid:b8a19f5afae13d3da38b359244dc0f560df03176
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:42
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1
repl_backlog_histlen:42
127.0.0.1:6379>

可见当前主从同步已建立。

二、哨兵模式搭建

1、创建3个哨兵实例

mkdir -p   /data/redis/redis_sentinel/
cd /data/redis/redis_sentinel/
mkdir sentinel26379 sentinel26380 sentinel26381
cd sentinel26379
 # 初始化配置文件如下 
vim  redis_sentinel_26379.conf
bind 0.0.0.0
port 26379
daemonize yes
dir "/data/redis/redis_sentinel/sentinel26379"
pidfile "/data/redis/redis_sentinel/sentinel26379/redis_sentinel26379.pid"
logfile "/data/redis/redis_sentinel/sentinel26379/redis_sentinel26379.log"
# Generated by CONFIG REWRITE
sentinel deny-scripts-reconfig yes
sentinel monitor testdb 127.0.0.1 6379 2
sentinel down-after-milliseconds testdb 5000
sentinel auth-pass testdb 123456
# 配置文件拷贝至另2个实例
cp  redis_sentinel_26379.conf   ../sentinel26380/redis_sentinel_26380.conf 
sed -i  "s#26379#26380#g"  ../sentinel26380/redis_sentinel_26380.conf 
cp  redis_sentinel_26379.conf   ../sentinel26381/redis_sentinel_26381.conf 
sed -i "s#26379#26381#g"  ../sentinel26381/redis_sentinel_26381.conf

配置文件主要参数说明：

参数名	说明
bind	绑定的可以访问的主机IP，0.0.0.0 代表不限制
port	哨兵实例的端口
sentinel monitor testdb 127.0.0.1 6379 1	testdb任意定义，哨兵集群名称，127.0.0.1 6379 redis实例主节点；1 代表当1个哨兵实例判断主库不可用则进行转移，生产环境节点数要配置多一点
sentinel down-after-milliseconds testdb 5000	testdb同上，down-after-milliseconds代表 master 最长响应时间，超过这个时间就主观判断它下线，5000 代表5000ms,即5s
sentinel auth-pass testdb 123456	123456 是redis实例的登录密码

2、启动哨兵实例

redis-sentinel  /data/redis/redis_sentinel/sentinel26379/redis_sentinel_26379.conf 
redis-sentinel  /data/redis/redis_sentinel/sentinel26380/redis_sentinel_26380.conf 
redis-sentinel  /data/redis/redis_sentinel/sentinel26381/redis_sentinel_26381.conf

启动后配置文件会自动新增如下红框中的内容。

登录哨兵实例查看。

redis-cli -p  26379

3、测试

测试将主节点down机。

redis-cli -p 6379 -a 123456  shutdown

再查看哨兵找那个的master结果，如下：

日志信息如下：

1895:X 29 Apr 23:57:31.778 # +sdown master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.779 # +odown master testdb 127.0.0.1 6379 #quorum 1/1
1895:X 29 Apr 23:57:31.779 # +new-epoch 1
1895:X 29 Apr 23:57:31.779 # +try-failover master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.795 # +vote-for-leader 4928b4d4dfd762cd50fa540b7a0903d2be3b0f95 1
1895:X 29 Apr 23:57:31.796 # +elected-leader master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.796 # +failover-state-select-slave master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.862 # +selected-slave slave 127.0.0.1:6380 127.0.0.1 6380 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.862 * +failover-state-send-slaveof-noone slave 127.0.0.1:6380 127.0.0.1 6380 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:31.991 * +failover-state-wait-promotion slave 127.0.0.1:6380 127.0.0.1 6380 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:32.223 # +promoted-slave slave 127.0.0.1:6380 127.0.0.1 6380 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:32.223 # +failover-state-reconf-slaves master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:32.273 * +slave-reconf-sent slave 127.0.0.1:6381 127.0.0.1 6381 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:33.231 * +slave-reconf-inprog slave 127.0.0.1:6381 127.0.0.1 6381 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:33.231 * +slave-reconf-done slave 127.0.0.1:6381 127.0.0.1 6381 @ testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:33.296 # +failover-end master testdb 127.0.0.1 6379
1895:X 29 Apr 23:57:33.296 # +switch-master testdb 127.0.0.1 6379 127.0.0.1 6380
1895:X 29 Apr 23:57:33.297 * +slave slave 127.0.0.1:6381 127.0.0.1 6381 @ testdb 127.0.0.1 6380
1895:X 29 Apr 23:57:33.297 * +slave slave 127.0.0.1:6379 127.0.0.1 6379 @ testdb 127.0.0.1 6380
1895:X 29 Apr 23:57:38.356 # +sdown slave 127.0.0.1:6379 127.0.0.1 6379 @ testdb 127.0.0.1 6380