Redis的复制Master-Slave

╰+哭是因爲堅強的太久メ 2022-09-26 14:56 48阅读 0赞

# Redis 的复制 Master/Slave #

## 一、是什么 ##

行话：也就是我们所说的主从复制，主机数据更新后根据配置和策略，  
自动同步到备机的master/slaver机制，Master以写为主，Slave以读为主

## 二、能干嘛 ##

*  读写分离
 *  容灾恢复

## 三、怎么玩 ##

### 3.1、配从（库）不配主（库） ###

### 3.2、从库配置 ###

命令：

slaveof 主库 IP 主库端口

*  每次与master断开之后，都需要重新连接，除非你配置进redis.conf文件
 *  info replication

### 3.3、修改配置文件细节操作 ###

*  拷贝多个 redis.conf 文件
 *  开启 daemonize yes
 *  Pid 文件名字
 *  指定端口
 *  Log 文件名字
 *  Dump.rdb 名字

redis6379.conf

daemonize yes
    pidfile /var/run/redis6379.pid
    prot 6379
    logfile "6379.log"
    dbfilename dump6379.rdb

redis6380.conf

daemonize yes
    pidfile /var/run/redis6380.pid
    prot 6380
    logfile "6380.log"
    dbfilename dump6380.rdb

redis6381.conf

daemonize yes
    pidfile /var/run/redis6381.pid
    prot 6381
    logfile "6381.log"
    dbfilename dump6381.rdb

### 3.4、常用3招 ###

#### 3.4.1、 一主二仆 ####

窗口 A

$ redis-server redis6379.conf 
    $ redis-cli -p 6379
    127.0.0.1:6379> ping
    PONG

窗口 B

$ redis-server redis6380.conf 
    $ redis-cli -p 6380
    127.0.0.1:6380> ping
    PONG

窗口 C

$ redis-server redis6381.conf 
    $ redis-cli -p 6381
    127.0.0.1:6381> ping
    PONG

窗口 A

127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0
    127.0.0.1:6379> set k1 v1
    OK
    127.0.0.1:6379> set k2 v2
    OK
    127.0.0.1:6379> set k3 v3
    OK
    127.0.0.1:6379> keys *
    1) "k1"
    2) "k2"
    3) "k3"

窗口 B

127.0.0.1:6380> SLAVEOF 127.0.0.1 6379
    OK
    127.0.0.1:6380> get k3
    "v3"

窗口 C

127.0.0.1:6381> SLAVEOF 127.0.0.1 6379
    OK
    127.0.0.1:6381> get k3
    "v3"

6380、6381 slaveof 之后

窗口 A

127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:2
    slave0:ip=127.0.0.1,port=6380,state=online,offset=337,lag=1
    slave1:ip=127.0.0.1,port=6381,state=online,offset=337,lag=1
    master_repl_offset:337
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:336

窗口 B

127.0.0.1:6380> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:up
    master_last_io_seconds_ago:8
    master_sync_in_progress:0
    slave_repl_offset:449
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

窗口 C

127.0.0.1:6381> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:up
    master_last_io_seconds_ago:2
    master_sync_in_progress:0
    slave_repl_offset:505
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

窗口 A

127.0.0.1:6379> set k5 v5
    OK

窗口 B 直接备份了

127.0.0.1:6380> get k5
    "v5"

窗口 C 直接备份了

127.0.0.1:6381> get k5
    "v5"

窗口 A

127.0.0.1:6379> set k6 v6
    OK

窗口 B

127.0.0.1:6380> set k6 v66
    (error) READONLY You can't write against a read only slave.

窗口 C

127.0.0.1:6381> set k6 v666
    (error) READONLY You can't write against a read only slave.

当主机 A 宕了，主机 B，主机 C 会不会上位

窗口 A

127.0.0.1:6379> SHUTDOWN
    not connected> exit

窗口 B , 角色还是 slave，连接状态是 down

127.0.0.1:6380> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:down
    master_last_io_seconds_ago:-1
    master_sync_in_progress:0
    slave_repl_offset:1230
    master_link_down_since_seconds:48
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0
    127.0.0.1:6380>

窗口 C ，角色还是 slave，连接状态是 down

127.0.0.1:6381> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:down
    master_last_io_seconds_ago:-1
    master_sync_in_progress:0
    slave_repl_offset:1230
    master_link_down_since_seconds:118
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

当主机 A 又上线了，从机依旧

窗口 A

$ redis-server redis6379.conf 
    $ redis-cli -p 6379
    127.0.0.1:6379> ping
    PONG
    127.0.0.1:6379> set k7 v7
    OK

窗口 B

127.0.0.1:6380> get k7
    "v7"

窗口 C

127.0.0.1:6381> get k7
    "v7"

当从机死了

窗口 B

127.0.0.1:6380> SHUTDOWN
    not connected> exit

窗口 A

127.0.0.1:6379> set k8 v8
    OK
    127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:1
    slave0:ip=127.0.0.1,port=6381,state=online,offset=348,lag=1
    master_repl_offset:348
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:347

窗口 C

127.0.0.1:6381> get k8
    "v8"

重新启动 B

$ redis-server redis6380.conf 
    $ redis-cli -p 6380
    127.0.0.1:6380> ping
    PONG
    127.0.0.1:6380> info replication
    # Replication
    role:master
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0
    127.0.0.1:6380> get k8
    (nil)

每次与master断开之后，都需要重新连接，除非你配置进redis.conf文件

重新连接 B

窗口 B

127.0.0.1:6380> SLAVEOF 127.0.0.1 6379
    OK
    127.0.0.1:6380> get k8
    "v8"

#### 3.4.2、 薪火相传 ####

*  上一个Slave可以是下一个slave的Master，Slave同样可以接收其他  
    slaves的连接和同步请求，那么该slave作为了链条中下一个的master,  
    可以有效减轻master的写压力
 *  中途变更转向:会清除之前的数据，重新建立拷贝最新的
 *  slaveof 新主库IP 新主库端口

窗口 C

127.0.0.1:6381> SLAVEOF 127.0.01 6380
    OK

窗口 A

127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:1
    slave0:ip=127.0.0.1,port=6380,state=online,offset=1370,lag=0
    master_repl_offset:1370
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:1369
    127.0.0.1:6379> set k9
    (error) ERR wrong number of arguments for 'set' command
    127.0.0.1:6379> set k9 v9
    OK

窗口 B

127.0.0.1:6380> get k9
    "v9"

窗口 C

127.0.0.1:6381> get k9
    "v9"

窗口 B , 角色还是 slave ， 连接了一个 slave

127.0.0.1:6380> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:up
    master_last_io_seconds_ago:3
    master_sync_in_progress:0
    slave_repl_offset:1384
    slave_priority:100
    slave_read_only:1
    connected_slaves:1
    slave0:ip=127.0.0.1,port=6381,state=online,offset=29,lag=0
    master_repl_offset:29
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:28

#### 3.4.3、 反客为主 ####

命令

SLAVEOF no one

使当前数据库停止与其他数据库的同步，转为主数据库

窗口 C (恢复一主二从)

127.0.0.1:6381> SLAVEOF 127.0.01 6379
    OK

窗口 A

127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:2
    slave0:ip=127.0.0.1,port=6381,state=online,offset=2206,lag=0
    slave1:ip=127.0.0.1,port=6380,state=online,offset=2206,lag=0
    master_repl_offset:2206
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:2205

窗口 B

127.0.0.1:6380> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6379
    master_link_status:up
    master_last_io_seconds_ago:10
    master_sync_in_progress:0
    slave_repl_offset:2206
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:865
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:864

窗口 C

127.0.0.1:6381> info replication
    # Replication
    role:slave
    master_host:127.0.01
    master_port:6379
    master_link_status:up
    master_last_io_seconds_ago:5
    master_sync_in_progress:0
    slave_repl_offset:2220
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

当主机 A 宕了

窗口 A

127.0.0.1:6379> SHUTDOWN
    not connected> exit

窗口 B

127.0.0.1:6380> info replication
    # Replication
    role:master
    connected_slaves:0
    master_repl_offset:1131
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:1130

窗口 C

127.0.0.1:6381> info replication
    # Replication
    role:slave
    master_host:127.0.01
    master_port:6379
    master_link_status:down
    master_last_io_seconds_ago:-1
    master_sync_in_progress:0
    slave_repl_offset:2332
    master_link_down_since_seconds:154
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0
    127.0.0.1:6381> SLAVEOF 127.0.0.1 6380
    OK
    127.0.0.1:6381> get k10
    "v10"

当 主机 A 重新上线

$ redis-server redis6379.conf 
    $ redis-cli -p 6379
    127.0.0.1:6379> ping
    PONG
    127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

## 四、复制原理 ##

*  slave启动成功连接到master后会发送一个sync命令
 *  Master接到命令启动后台的存盘进程，同时收集所有接收到的用于修改数据集命令，  
    在后台进程执行完毕之后，master将传送整个数据文件到slave,以完成一次完全同步
 *  全量复制：而slave服务在接收到数据库文件数据后，将其存盘并加载到内存中。
 *  增量复制：Master继续将新的所有收集到的修改命令依次传给slave,完成同步
 *  但是只要是重新连接master,一次完全同步（全量复制)将被自动执行

## 五、哨兵模式 （sentinel） ##

### 5.1、是什么 ###

反客为主的自动版，能够后台监控主机是否故障，如果故障了根据投票数自动将从库转换为主库

### 5.2、怎么玩 ###

*  调整结构，6379带着80、81
 *  自定义的/myredis目录下新建sentinel.conf文件，名字绝不能错
 *  配置哨兵,填写内容  
    
    
     *  sentinel monitor 被监控数据库名字(自己起名字) 127.0.0.1 6379 1
     *  上面最后一个数字1，表示主机挂掉后 salve 投票看让谁接替成为主机，得票数多少后成为主机
 *  启动哨兵  
    
    
     *  redis-sentinel /myredis/sentinel.conf
     *  上述目录依照各自的实际情况配置，可能目录不同
 *  正常主从演示
 *  原有的master挂了
 *  投票新选
 *  重新主从继续开工,info replication查查看
 *  问题：如果之前的master重启回来，会不会双master冲突？

窗口 B 恢复一主二从

127.0.0.1:6380> SLAVEOF 127.0.0.1 6379
    OK

窗口 C

127.0.0.1:6381> SLAVEOF 127.0.0.1 6379
    OK

窗口 A

127.0.0.1:6379> info replication
    # Replication
    role:master
    connected_slaves:2
    slave0:ip=127.0.0.1,port=6380,state=online,offset=141,lag=0
    slave1:ip=127.0.0.1,port=6381,state=online,offset=141,lag=1
    master_repl_offset:141
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:140

sentinel.conf

sentinel monitor host6379 127.0.0.1 6379 1

窗口 D ,开启哨兵巡逻

$ redis-sentinel sentinel.conf 
    9133:X 03 Mar 11:17:50.266 * Increased maximum number of open files to 10032 (it was originally set to 256).
                    _._                                                  
               _.-``__ ''-._                                             
          _.-``    `. `_.  ''-._           Redis 3.2.6 (00000000/0) 64 bit
      .-`` .-```. ```\/    _.,_ ''-._                                   
     (    ' , .-` | `, ) Running in sentinel mode |`-._`-...-` __...-.``-._|'` _.-'| Port: 26379 | `-._   `._ / _.-' | PID: 19133 `-._    `-._ `-./  _.-' _.-'                                   
     |`-._`-._    `-.__.-' _.-'_.-'| | `-._`-._ _.-'_.-' | http://redis.io `-._    `-._`-.__.-'_.-'    _.-' |`-._`-._ `-.__.-'    _.-'_.-'|                                  
     |    `-._`-._        _.-'_.-'    |                                  
      `-._ `-._`-.__.-'_.-' _.-' `-._    `-.__.-' _.-' `-._        _.-' `-.__.-'                                               
    
    19133:X 03 Mar 11:17:50.268 # Sentinel ID is a102738f97ae0492bf807bd262553bc011a19e6c
    19133:X 03 Mar 11:17:50.268 # +monitor master host6379 127.0.0.1 6379 quorum 1
    19133:X 03 Mar 11:17:50.269 * +slave slave 127.0.0.1:6380 127.0.0.1 6380 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:17:50.269 * +slave slave 127.0.0.1:6381 127.0.0.1 6381 @ host6379 127.0.0.1 6379

主机 A 宕了

窗口 A

127.0.0.1:6379> SHUTDOWN
    not connected> exit

窗口 D ,哨兵投票选取新主机

19133:X 03 Mar 11:20:37.775 # +sdown master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.775 # +odown master host6379 127.0.0.1 6379 #quorum 1/1
    19133:X 03 Mar 11:20:37.775 # +new-epoch 1
    19133:X 03 Mar 11:20:37.775 # +try-failover master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.777 # +vote-for-leader a102738f97ae0492bf807bd262553bc011a19e6c 1
    19133:X 03 Mar 11:20:37.777 # +elected-leader master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.777 # +failover-state-select-slave master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.853 # +selected-slave slave 127.0.0.1:6380 127.0.0.1 6380 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.853 * +failover-state-send-slaveof-noone slave 127.0.0.1:6380 127.0.0.1 6380 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:37.932 * +failover-state-wait-promotion slave 127.0.0.1:6380 127.0.0.1 6380 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:38.900 # +promoted-slave slave 127.0.0.1:6380 127.0.0.1 6380 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:38.900 # +failover-state-reconf-slaves master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:38.955 * +slave-reconf-sent slave 127.0.0.1:6381 127.0.0.1 6381 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:39.923 * +slave-reconf-inprog slave 127.0.0.1:6381 127.0.0.1 6381 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:39.923 * +slave-reconf-done slave 127.0.0.1:6381 127.0.0.1 6381 @ host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:39.980 # +failover-end master host6379 127.0.0.1 6379
    19133:X 03 Mar 11:20:39.980 # +switch-master host6379 127.0.0.1 6379 127.0.0.1 6380
    19133:X 03 Mar 11:20:39.981 * +slave slave 127.0.0.1:6381 127.0.0.1 6381 @ host6379 127.0.0.1 6380
    19133:X 03 Mar 11:20:39.981 * +slave slave 127.0.0.1:6379 127.0.0.1 6379 @ host6379 127.0.0.1 6380

窗口 B

127.0.0.1:6380> info replication
    # Replication
    role:master
    connected_slaves:1
    slave0:ip=127.0.0.1,port=6381,state=online,offset=26333,lag=1
    master_repl_offset:26333
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:26332
    127.0.0.1:6380>

窗口 C

127.0.0.1:6381> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6380
    master_link_status:up
    master_last_io_seconds_ago:0
    master_sync_in_progress:0
    slave_repl_offset:27705
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

窗口 B

127.0.0.1:6380> set k10 v10
    OK

窗口 C

127.0.0.1:6381> get k10
    "v10"

当主机 A 重新上线,哨兵监控到后，老领导变成从机了

$ redis-server redis6379.conf 
    $ redis-cli -p 6379
    127.0.0.1:6379> info replication
    # Replication
    role:slave
    master_host:127.0.0.1
    master_port:6380
    master_link_status:up
    master_last_io_seconds_ago:0
    master_sync_in_progress:0
    slave_repl_offset:40793
    slave_priority:100
    slave_read_only:1
    connected_slaves:0
    master_repl_offset:0
    repl_backlog_active:0
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:0
    repl_backlog_histlen:0

窗口 B

127.0.0.1:6380> info replication
    # Replication
    role:master
    connected_slaves:2
    slave0:ip=127.0.0.1,port=6381,state=online,offset=45413,lag=0
    slave1:ip=127.0.0.1,port=6379,state=online,offset=45413,lag=0
    master_repl_offset:45413
    repl_backlog_active:1
    repl_backlog_size:1048576
    repl_backlog_first_byte_offset:2
    repl_backlog_histlen:45412

### 5.3、一组 sentinel 能同时监控多个 Master ###

## 六、复制的缺点 ##

复制延迟

由于所有的写操作都是先在Master上操作，然后同步更新到Slave上，所以从Master同步到Slave机器有一定的延迟，当系统很繁忙的时候，延迟问题会更加严重，Slave机器数量的增加也会使这个问题更加严重。