本文共 10091 字，大约阅读时间需要 33 分钟。

一、redis集群之主从复制

1.1、集群介绍

在主从模式的Redis系统中，从数据库在整个系统中起到了数据冗余备份和读写分离的作用。

1.2、搭建过程

1、先给每个节点安装redis

[root@localhost opt]# tar xzvf redis-5.0.7.tar.gz -C /opt/[root@localhost opt]# cd /opt/redis-5.0.7/[root@localhost redis-5.0.7]# make        [root@localhost redis-5.0.7]#  make PREFIX=/usr/local/redis install    ####如果安装过程中，更改安装路径可以用make PRRFIX=安装路径 install//设置Redis相关配置文件[root@localhost redis-5.0.7]# ln -s /usr/local/redis/bin/* /usr/local/bin/[root@localhost redis-5.0.7]# cd utils/[root@localhost utils]# ./install_server.sh Welcome to the redis service installerThis script will help you easily set up a running redis serverPlease select the redis port for this instance: [6379] Selecting default: 6379Please select the redis config file name [/etc/redis/6379.conf] Selected default - /etc/redis/6379.confPlease select the redis log file name [/var/log/redis_6379.log] Selected default - /var/log/redis_6379.logPlease select the data directory for this instance [/var/lib/redis/6379] Selected default - /var/lib/redis/6379Please select the redis executable path [/usr/local/bin/redis-server] /usr/local/bin/redis-server  #### 未做软连接，可手动输入Selected config:Port           : 6379                                        ####端口号Config file    : /etc/redis/6379.conf                        ####默认配置文件Log file       : /var/log/redis_6379.log                     ####日志文件Data dir       : /var/lib/redis/6379                         ####数据目录Executable     : /usr/local/bin/redis-server                 ####执行命令Cli Executable : /usr/local/bin/redis-cli                    ####客户端命令Is this ok? Then press ENTER to go on or Ctrl-C to abort.Copied /tmp/6379.conf => /etc/init.d/redis_6379Installing service...Successfully added to chkconfig!Successfully added to runlevels 345!Starting Redis server...Installation successful!

2、修改主、从节点的配置文件

主节点：vi /etc//redis/6379.conf  70/bind 0.0.0.0在研发、测试环境可以考虑bind 0.0.0.0，线上生产环境建议绑定从节点服务器的IP地址sed -i '69s/127.0.0.1/0.0.0.0/' /etc/redis/6379.conf 137行 daemonize no     #开启守护进程yessed -i '136s/no/yes/' /etc//redis/6379.conf 172行 logfile ""   改成 logfile /var/log/redis_6379.logsed -i '171s/""/\/var\/log\/redis_6379.log/' /etc/redis/6379.conf 264行 dir ./  改成  dir /var/lib/redis/6379   #本地数据库的存放目录sed -i '263s/.\//\/var\/lib\/redis\/6379/' /etc/redis/6379.conf 700/appendonly no    #启用AOF持久化sed -i '699s/no/yes/' /etc/redis/6379.conf

从节点：vi /etc//redis/6379.conf 67/bind 0.0.0.0在研发、测试环境可以考虑bind 0.0.0.0，线上生产环境建议绑定从节点服务器的IP地址sed -i '69s/127.0.0.1/0.0.0.0/' /etc/redis/6379.conf 137行 daemonize no     #开启守护进程yessed -i '136s/no/yes/' /etc/redis/6379.conf 172行 logfile ""   改成 logfile /var/log/redis_6379.logsed -i '171s/""/\/var\/log\/redis_6379.log/' /etc/redis/6379.conf 264行 dir ./  改成  dir /var/lib/redis/6379sed -i '263s/.\//\/var\/lib\/redis\/6379/' /etc/redis/6379.conf 700/appendonly no    #启用AOF持久化sed -i '699s/no/yes/' /etc//redis/6379.conf 287/# replicaof 
    
    
      改成 replicaof 10.0.0.10 6379sed -i '286creplicaof 10.0.0.10 6379' /etc/redis/6379.conf
    
   

3、启动服务，先启动master redis，然后依次启动slave

#关闭防火墙systemctl stop firewalld.servicesetenforce 0#重启服务/etc/init.d/redis_6379 restartStopping ...Waiting for Redis to shutdown ...Redis stoppedStarting Redis server...#检查端口netstat -anpt | grep redistcp        0      0 0.0.0.0:6379            0.0.0.0:*               LISTEN      9923/redis-server 0

4、 查看日志文件内容，可以看到两个slave节点连接成功的信息！

tail -f /var/log/redis_6379.log

[root@server1 ~]# tail -f /var/log/redis_6379.log22035:M 12 Nov 2020 11:01:40.860 * Background saving terminated with success22035:M 12 Nov 2020 11:01:40.860 * Synchronization with replica 10.0.0.30:6379 succeeded22035:M 12 Nov 2020 11:01:41.112 * Replica 10.0.0.20:6379 asks for synchronization22035:M 12 Nov 2020 11:01:41.113 * Full resync requested by replica 10.0.0.20:637922035:M 12 Nov 2020 11:01:41.113 * Starting BGSAVE for SYNC with target: disk22035:M 12 Nov 2020 11:01:41.114 * Background saving started by pid 2214822148:C 12 Nov 2020 11:01:41.119 * DB saved on disk22148:C 12 Nov 2020 11:01:41.120 * RDB: 0 MB of memory used by copy-on-write22035:M 12 Nov 2020 11:01:41.163 * Background saving terminated with success22035:M 12 Nov 2020 11:01:41.163 * Synchronization with replica 10.0.0.20:6379 succeeded

也可以通过登录数据库，查看节点角色

[root@server1 ~]# redis-cli127.0.0.1:6379> info replication# Replicationrole:masterconnected_slaves:2slave0:ip=10.0.0.30,port=6379,state=online,offset=1498,lag=1slave1:ip=10.0.0.20,port=6379,state=online,offset=1498,lag=0master_replid:388136944386ad3b8b7a1b640fedb19cf9897fa8master_replid2:0000000000000000000000000000000000000000master_repl_offset:1498second_repl_offset:-1repl_backlog_active:1repl_backlog_size:1048576repl_backlog_first_byte_offset:1repl_backlog_histlen:1498

1.3、效果验证

进入主节点的数据库中，创建键值对，在从节点中可以进行查询，但是不能在从节点创建新的键值对。

[root@localhost redis]# redis-cli127.0.0.1:6379> set name lili	   #创建一个键值对OK127.0.0.1:6379> keys *	  #查看所有键1) "name"127.0.0.1:6379> get name	    #查看name键的值"lili"127.0.0.1:6379> exit

从节点服务器：

[root@localhost ~]# redis-cli 	 127.0.0.1:6379> get name	    #查看name键的值"lili"127.0.0.1:6379> exit

二、redis集群之哨兵模式

2.1、部署过程

redis2.8版本之前是主从复制模式，但是当数据库遇到异常中断服务后，我们只能通过手动的方式选择一个从数据库来升格为主数据库，显然这种方式很麻烦需要人工介入，这时通过哨兵模式可以实现自动化的系统监控和故障恢复。

哨兵模式是在主从复制的基础上加入监控机制，来实现自动化的系统监控和故障恢复功能。

哨兵集群的结构图：

vi /opt/redis-5.0.7/sentinel.conf17/protected-mode no    #关闭保护模式sed -i '17s/^#//' /opt/redis-5.0.7/sentinel.confsed -i '17s/^ //' /opt/redis-5.0.7/sentinel.conf21/port 26379     #无需修改,sentinel的端口号26/daemonize yes    指定sentinel为后台启动sed -i '26s/no/yes/' /opt/redis-5.0.7/sentinel.conf36/logfile /var/log/sentinel.log   #日志存放路径sed -i '36s/""/\/var\/log\/sentinel.log/' /opt/redis-5.0.7/sentinel.conf65/dir /tmp 改成 dir "/var/lib/redis/6379"    #数据库存放路径sed -i '65s/tmp/var\/lib\/redis\/6379/' /opt/redis-5.0.7/sentinel.conf84/sentinel monitor mymaster 10.0.0.10 6379 2sed -i '84s/127.0.0.1/192.168.100.128/' /opt/redis-5.0.7/sentinel.conf//sentinel monitor mymaster [主服务器节点IP] [主服务器节点port] [几个哨兵检测到主节点失联数量]  -- 至少有几个哨兵检测到主服务器故障了，才进行故障迁移 113/sentinel down-after-milliseconds mymaster 3000     #原来是30000，为了尽快看到效果，将时间缩短  sed -i '113s/30000/3000/' /opt/redis-5.0.7/sentinel.conf146/sentinel failover-timeout mymaster 120000   #原来是180000sed -i '146s/180000/120000/' /opt/redis-5.0.7/sentinel.conf

slave节点：

vi /opt/redis-5.0.7/sentinel.conf17/protected-mode no    #关闭保护模式sed -i '17s/^#//' /opt/redis-5.0.7/sentinel.confsed -i '17s/^ //' /opt/redis-5.0.7/sentinel.conf26/daemonize yes    指定sentinel为后台启动sed -i '26s/no/yes/' /opt/redis-5.0.7/sentinel.conf36/logfile /var/log/sentinel.log   #日志存放路径65/dir /tmp 改成 dir "/var/lib/redis/6379"    #数据库存放目录84/sentinel monitor mymaster 10.0.0.10 6379 2    #填写主节点的地址113/sentinel down-after-milliseconds mymaster 3000     #原来是30000，为了尽快看到效果，将时间缩短  sed -i '113s/30000/3000/' /opt/redis-5.0.7/sentinel.conf146/sentinel failover-timeout mymaster 120000   #原来是180000sed -i '146s/180000/120000/' /opt/redis-5.0.7/sentinel.conf

3、启动哨兵服务

#放在后台运行[root@server1 ~]# redis-sentinel /opt/redis-5.0.7/sentinel.conf &[1] 22804

查看哨兵信息，里面有master节点的IP地址

[root@server1 ~]# redis-cli -p 26379 INFO Sentinel# Sentinelsentinel_masters:1sentinel_tilt:0sentinel_running_scripts:0sentinel_scripts_queue_length:0sentinel_simulate_failure_flags:0master0:name=mymaster,status=ok,address=10.0.0.10:6379,slaves=2,sentinels=3[1]+  完成                  redis-sentinel /opt/redis-5.0.7/sentinel.conf

2.2、故障模拟

关闭主节点的redis-server服务

ps -ef | grep redis    #查看redis-server进程数kill -9 进程数   #关闭redis-server#查看日志tail -99f /var/log/sentinel.log 看到switch-master mymaster 10.0.0.10 6379 变为 10.0.0.30 6379，即主节点down掉，从节点顶替主节点。#查看哨兵信息，可以看到master节点的IP地址改变[root@server1 ~]# redis-cli -p 26379 INFO Sentinel# Sentinelsentinel_masters:1sentinel_tilt:0sentinel_running_scripts:0sentinel_scripts_queue_length:0sentinel_simulate_failure_flags:0master0:name=mymaster,status=ok,address=10.0.0.30:6379,slaves=2,sentinels=3

22804:X 12 Nov 2020 11:48:23.846 # Configuration loaded22805:X 12 Nov 2020 11:48:23.847 * Increased maximum number of open files to 10032 (it was originally set to 1024).22805:X 12 Nov 2020 11:48:23.848 * Running mode=sentinel, port=26379.22805:X 12 Nov 2020 11:48:23.848 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.22805:X 12 Nov 2020 11:48:23.850 # Sentinel ID is 6adf7655f16d6cfb58b456be98c9a732509ceac322805:X 12 Nov 2020 11:48:23.850 # +monitor master mymaster 10.0.0.10 6379 quorum 222805:X 12 Nov 2020 11:48:23.851 * +slave slave 10.0.0.30:6379 10.0.0.30 6379 @ mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:48:23.852 * +slave slave 10.0.0.20:6379 10.0.0.20 6379 @ mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:48:37.908 * +sentinel sentinel a070582084975c174cfafbc69eb01435cca39415 10.0.0.20 26379 @ mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:48:42.632 * +sentinel sentinel 60e5a16fdef37b9f8b204a29a11a2bce4d7dfdab 10.0.0.30 26379 @ mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:56:28.410 # +sdown master mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:56:28.449 # +new-epoch 122805:X 12 Nov 2020 11:56:28.451 # +vote-for-leader a070582084975c174cfafbc69eb01435cca39415 122805:X 12 Nov 2020 11:56:28.477 # +odown master mymaster 10.0.0.10 6379 #quorum 3/222805:X 12 Nov 2020 11:56:28.477 # Next failover delay: I will not start a failover before Thu Nov 12 12:02:28 202022805:X 12 Nov 2020 11:56:28.834 # +config-update-from sentinel a070582084975c174cfafbc69eb01435cca39415 10.0.0.20 26379 @ mymaster 10.0.0.10 637922805:X 12 Nov 2020 11:56:28.834 # +switch-master mymaster 10.0.0.10 6379 10.0.0.30 637922805:X 12 Nov 2020 11:56:28.835 * +slave slave 10.0.0.20:6379 10.0.0.20 6379 @ mymaster 10.0.0.30 637922805:X 12 Nov 2020 11:56:28.835 * +slave slave 10.0.0.10:6379 10.0.0.10 6379 @ mymaster 10.0.0.30 637922805:X 12 Nov 2020 11:56:58.844 # +sdown slave 10.0.0.10:6379 10.0.0.10 6379 @ mymaster 10.0.0.30 6379

转载地址：http://lekgf.baihongyu.com/