redis sentinel集群安装命令及问题处理

参考官网：redis from source

测试机器（sentine+redis)

redis server端口：6380

密码：*********

10.10.10.3 （master）

10.10.10.2 (slave)

10.10.10.1 (slave)

sentinel端口：26380

sentinel列表：10.10.10.3 10.10.10.2 10.10.10.1

安装步骤

1. 下载

   get https://download.redis.io/redis-stable.tar.gz
   

2. 编译

   tar -xzvf redis-stable.tar.gz
   cd redis-stable
   make
   

3. 安装

   # To install these binaries in /usr/local/bin
   make install
   

4. 启动测试

   redis-server
   

5. 主从配置

   cd /data/apps/redis-6.2.6
   mkdir config
   mkdir logs
   vim config/redis-6380.conf
   

   master配置文件

   port 6380
   pidfile /data/apps/redis-6.2.6/redis_6380.pid 
   requirepass ********    
   protected-mode  no   
   daemonize  yes
   bind 0.0.0.0
   #从配置加上以下
   #replicaof 10.10.10.3 6380
   #masterauth 	**********
   

   持久化策略配置

   #RDB的默认策略
   #表示在 3600 秒内，至少更新了 1 条数据，Redis 自动触发 BGSAVE 命令，将数据保存到硬盘。
   save 3600 1 
   #表示在 300 秒内，至少更新了 100 条数据，Redis 自动触 BGSAVE 命令，将数据保存到硬盘。
   save 300 100
   #表示 60 秒内，至少更新了 10000 条数据，Redis 自动触发 BGSAVE 命令，将数据保存到硬盘。
   save 60 10000 
   #开启aof，默认no
   appendonly yes
   

6. 启动服务

# 分别启动主从三台机器
/data/apps/redis-6.2.6/src/redis-server /data/apps/redis-6.2.6/config/redis-6380.conf &

7. 验证主从同步

   #master 
   127.0.0.1:6380> info
   # Replication
   role:master
   connected_slaves:2
   slave0:ip=10.10.10.2,port=6380,state=online,offset=84,lag=0
   slave1:ip=10.0.10.1,port=6380,state=online,offset=84,lag=0
   master_failover_state:no-failover
   127.0.0.1:6380> set test test
   OK
   #slave
   127.0.0.1:6380> get test
   "test"
   

配置Sentinel

1. 新建配置文件

   cd /data/apps/redis-6.2.6
   cp sentinel.conf config/sentinel-26380.conf
   

   配置文件内容

   sentinel三台节点增加以下配置

   scp sentinel-26380.conf root@node2:/data/apps/redis-6.2.6/config

   scp sentinel-26380.conf root@node3:/data/apps/redis-6.2.6/config

   port 26380
   daemonize yes
   pidfile "/data/apps/redis-6.2.6/redis-sentinel-26380.pid"
   logfile "/data/apps/redis-6.2.6/sentinel-26380.log"
   sentinel monitor mymaster 10.10.10.3 6380 2
   sentinel auth-pass mymaster ******
   sentinel down-after-milliseconds mymaster 30000
   sentinel failover-timeout mymaster 180000
   

   quorum指明当有多少个sentinel认为一个master失效时(值一般为：sentinel总数/2 + 1)，

   master才算真正失效

2. 启动

   /data/apps/redis-6.2.6/src/redis-sentinel /data/apps/redis-6.2.6/config/sentinel-26380.conf &
   

3. 测试验证

   • sentinel主从节点配置文件自动生成以下内容

     # Generated by CONFIG REWRITE
     protected-mode no
     user default on nopass sanitize-payload ~* &* +@all
     dir "/data/apps/redis-6.2.6/config"
     sentinel myid effab961c271f376dc9756f48fe7eb65344bb044
     sentinel config-epoch mymaster 0
     sentinel leader-epoch mymaster 0
     sentinel current-epoch 0
     sentinel known-sentinel mymaster 10.10.10.3 26380 06ff1d8a30bcd1d8c49f4c0c012a7a49c23cc2ce
     sentinel known-replica mymaster 10.10.10.1 6380
     sentinel known-sentinel mymaster 10.10.10.2 26380 926466d6701e9c412684b06eb971720ccb03486a
     sentinel known-replica mymaster 10.10.10.2 6380
     

   • 查看info

     [root@]# redis-cli -p 26380
     127.0.0.1:26380> info
     # Sentinel
     sentinel_masters:1
     sentinel_tilt:0
     sentinel_running_scripts:0
     sentinel_scripts_queue_length:0
     sentinel_simulate_failure_flags:0
     master0:name=mymaster,status=ok,address=10.10.10.3:6380,slaves=2,sentinels=3
     

   • 验证redis master宕机，主从自动切换

     #在redismaster 上执行stop redis,30s后查看
     127.0.0.1:26380> info sentinel
     # Sentinel
     sentinel_masters:1
     sentinel_tilt:0
     sentinel_running_scripts:0
     sentinel_scripts_queue_length:0
     sentinel_simulate_failure_flags:0
     master0:name=mymaster,status=ok,address=10.10.10.2:6380,slaves=2,sentinels=3
     #redis master已经切换为10.10.10.2为主，查看master info
     127.0.0.1:6380> info replication
     # Replication
     role:master
     connected_slaves:1
     slave0:ip=10.10.10.1,port=6380,state=online,offset=139145,lag=0
     master_failover_state:no-failover
     master_replid:15a89a2d2953d142a8e3f1e07cd849190ad3eada
     master_replid2:e8ddb137361138d8048d91f988758e01a366d02b
     master_repl_offset:139159
     second_repl_offset:111543
     repl_backlog_active:1
     repl_backlog_size:1048576
     repl_backlog_first_byte_offset:1
     repl_backlog_histlen:139159
     
     

配置supervisor管理

注意：需要设置redis和sentinel为前台启动daemonize no

1. vim /etc/supervisor.d/redis.conf

   [program:redis]
   command=/data/apps/redis-6.2.6/src/redis-server /data/apps/redis-6.2.6/config/redis-6380.conf
   priority=100
   process_name=%(program_name)s
   user=root
   autostart=true
   autorestart=true
   startsecs=3
   stdout_logfile=/data/apps/redis-6.2.6/logs/redis_stdout.log
   stdout_logfile_maxbytes=100MB
   stdout_logfile_backups=5
   stderr_logfile=/data/apps/redis-6.2.6/logs/redis_std_error.log
   stderr_logfile_maxbytes=100MB
   stderr_logfile_backups=5
   

2. vim /etc/supervisor.d/redis_sentinel.conf

   [program:redis_sentinel]
   command=/data/apps/redis-6.2.6/src/redis-sentinel /data/apps/redis-6.2.6/config/sentinel-26380.conf
   priority=100
   process_name=%(program_name)s
   user=root
   autostart=true
   autorestart=true
   startsecs=3
   stdout_logfile=/data/apps/redis-6.2.6/logs/sentinel_stdout.log
   stdout_logfile_maxbytes=100MB
   stdout_logfile_backups=5
   stderr_logfile=/data/apps/redis-6.2.6/logs/sentinel_std_error.log
   stderr_logfile_maxbytes=100MB
   stderr_logfile_backups=5
   

错误记录

• make 编译报错

   fatal error: systemd/sd-daemon.h: No such file or directory
  

  yum install systemd-devel
  

• make test 报错

  You need tcl 8.5 or newer in order to run the Redis test
  

  解决：

  yum install tcl.x86_64
  

• supervisor配置启动失败

  解决：

  #修改启动方式为前台启动
  daemonize no
  

参考文献

安装：https://redis.io/docs/getting-started/installation/install-redis-from-source/

搭建集群：https://www.cnblogs.com/kevingrace/p/9004460.html