Redis 迈向高可用之路：哨兵与 Cluster 解析

刘大帅ps

于 2024-08-29 09:00:00 发布

阅读量1.7k

点赞数 54

文章标签： redis 数据库缓存运维网络 linux sentinel

本文链接：https://blog.csdn.net/weixin_64373152/article/details/141650379

版权

6.2. sentinel中的三个定时任务

6.3. 哨兵的实验过程

6.4. 在整个架构中可能会出现的问题

七. Redis Cluster（无中心化设计）

7.1. Redis Cluster 工作原理

7.2. Redis Cluster特点如下

7.3. Redis cluster 主从架构

7.4. 创建redis cluster的前提

7.5. 部署redis cluster

7.6. 集群扩容

7.7. clsuter集群维护

五. Redis 主从复制

5.1. 配置主从同步

#修改mastser节点的配置文件
[root@redis-node1 & node2 & node3 ~]# vim /etc/redis/6379.conf
protected-mode no #关闭protected模式
[root@redis-node1 &node2 & node3 ~]# /etc/init.d/redis_6379 restart
Stopping ...
Redis stopped
Starting Redis server...

#配置slave节点
[root@redis-node2 & 3 ~]# vim /etc/redis/6379.conf
replicaof 172.25.254.100 6379
[root@redis-node2 & 3 ~]# /etc/init.d/redis_6379 restart
Stopping ...
Waiting for Redis to shutdown ...
Redis stopped
Starting Redis server...

#测试效果
#在mastser节点
[root@node1 ~]# redis-cli
127.0.0.1:6379> info replication
# Replication
role:master
connected_slaves:2
slave0:ip=172.25.254.20,port=6379,state=online,offset=714,lag=1
slave1:ip=172.25.254.30,port=6379,state=online,offset=714,lag=1
master_failover_state:no-failover
master_replid:23035f194cb43a56b8a24e0c92fc8b8ce17c1f1d
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:714
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1
repl_backlog_histlen:714
127.0.0.1:6379> set name zhuzhuxia
OK
127.0.0.1:6379> get name
"zhuzhuxia"

#在slave节点查看
[root@node2 ~]# redis-cli 
127.0.0.1:6379> get name
"zhuzhuxia"

5.2. 主从同步过程

slave节点发送同步亲求到master节点
slave节点通过master节点的认证开始进行同步
master节点会开启bgsave进程发送内存rbd到slave节点，在此过程中是异步操作，也就是说
master节点仍然可以进行写入动作
slave节点收到rdb后首先清空自己的所有数据
slave节点加载rdb并进行数据恢复
在master和slave同步过程中master还会开启新的bgsave进程把没有同步的数据进行缓存
然后通过自有的replactionfeedslave函数把未通过内存快照发动到slave的数据一条一条写入到slave中

六. Redis的哨兵（高可用）

实验环境：用主两从来实现Redis的高可用架构

6.1. Redis哨兵

Sentinel 进程是用于监控redis集群中Master主服务器工作的状态，在Master主服务器发生故障的时候，可以实现Master和Slave服务器的切换，保证系统的高可用，此功能在redis2.6+的版本已引用，Redis的哨兵模式到了2.8版本之后就稳定了下来。一般在生产环境也建议使用Redis的2.8版本的以后版本。
每个哨兵(Sentinel)进程会向其它哨兵(Sentinel)、Master、Slave定时发送消息，以确认对方是否”活”着，如果发现对方在指定配置时间(此项可配置)内未得到回应，则暂时认为对方已离线，也就是所谓的”主观认为宕机” (主观:是每个成员都具有的独自的而且可能相同也可能不同的意识)，英文名称：Subjective Down，简称SDOWN
有主观宕机，对应的有客观宕机。当“哨兵群”中的多数Sentinel进程在对Master主服务器做出SDOWN 的判断，并且通过 SENTINEL is-master-down-by-addr 命令互相交流之后，得出的Master Server下线判断，这种方式就是“客观宕机”(客观:是不依赖于某种意识而已经实际存在的一切事物)，英文名称是：Objectively Down，简称 ODOWN
通过一定的vote算法，从剩下的slave从服务器节点中，选一台提升为Master服务器节点，然后自动修改相关配置，并开启故障转（failover）
Sentinel 机制可以解决master和slave角色的自动切换问题，但单个 Master 的性能瓶颈问题无法解决,类似于MySQL中的MHA功能
Redis Sentinel中的Sentinel节点个数应该为大于等于3且最好为奇数

6.2. sentinel中的三个定时任务

每10秒每个sentinel对master和slave执行info -- 发现slave节点 -- 确认主从关系
每2秒每个sentinel通过master节点的channel交换信息(pub/sub) -- 通过sentinel__: hello频道交互 -- 交互对节点的“看法”和自身信息
每1秒每个sentinel对其他sentinel和redis执行pi

6.3. 哨兵的实验过程

在所有阶段中关闭 protected-mode no

#在master节点中
#编辑配置文件
[root@node1 ~]# cd redis-7.4.0/
[root@node1 redis-7.4.0]# cp sentinel.conf /etc/redis/
[root@node1 redis-7.4.0]# vim /etc/redis/sentinel.conf
protected-mode no #关闭保护模式
port 26379 #监听端口
daemonize no #进入不打如后台
pidfile /var/run/redis-sentinel.pid   #sentinel进程pid文件
loglevel notice #日志级别
sentinel monitor mymaster 172.25.254.10 6379 2    #修改这里    #创建sentinel监控监控master主机，2表示必须得到2票
sentinel down-after-milliseconds mymaster 10000    #master中断时长，10秒连不上视为master下线
sentinel parallel-syncs mymaster 1 #发生故障转移后，同时开始同步新master数据的slave数量
sentinel failover-timeout mymaster 180000 #整个故障切换的超时时间为3分钟

####复制配置文件到其他阶段
[root@node1 ~]# scp /etc/redis/sentinel.conf root@172.25.254.20:/etc/redis/
root@172.25.254.20's password: 
sentinel.conf                     100%   14KB  12.8MB/s   00:00    
[root@node1 ~]# scp /etc/redis/sentinel.conf root@172.25.254.30:/etc/redis/
root@172.25.254.30's password: 
sentinel.conf                     100%   14KB   6.8MB/s   00:00 

#启动服务
#三台机子都需要启动
[root@node1 redis-7.4.0]# redis-sentinel /etc/redis/sentinel.conf
39319:X 05 Aug 2024 20:53:16.807 # WARNING Memory overcommit must be enabled!
Without it, a background save or replication may fail under low memory condition.
Being disabled, it can also cause failures without low memory condition, see
https://github.com/jemalloc/jemalloc/issues/1328. To fix this issue add
'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command
'sysctl vm.overcommit_memory=1' for this to take effect.
39319:X 05 Aug 2024 20:53:16.807 * oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
39319:X 05 Aug 2024 20:53:16.807 * Redis version=7.4.0, bits=64, commit=00000000,
modified=0, pid=39319, just started
39319:X 05 Aug 2024 20:53:16.807 * Configuration loaded
39319:X 05 Aug 2024 20:53:16.808 * Increased maximum number of open files to
10032 (it was originally set to 1024).
39319:X 05 Aug 2024 20:53:16.808 * monotonic clock: POSIX clock_gettime
_._
_.-``__ ''-._
_.-`` `. `_. ''-._ Redis Community Edition
.-`` .-```. ```\/ _.,_ ''-._ 7.4.0 (00000000/0) 64 bit
( ' , .-` | `, ) Running in sentinel mode
|`-._`-...-` __...-.``-._|'` _.-'| Port: 26379
| `-._ `._ / _.-' | PID: 39319
`-._ `-._ `-./ _.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' | https://redis.io
`-._ `-._`-.__.-'_.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' |
`-._ `-._`-.__.-'_.-' _.-'
`-._ `-.__.-' _.-'
`-._ _.-'
`-.__.-'
39319:X 05 Aug 2024 20:53:16.810 * Sentinel new configuration saved on disk
39319:X 05 Aug 2024 20:53:16.810 * Sentinel ID is
e568add863fd7c132e03f7a6ce2c5ef367d3bdae
39319:X 05 Aug 2024 20:53:16.810 # +monitor master mymaster 172.25.254.100 6379
quorum 2
39319:X 05 Aug 2024 20:53:16.811 * +slave slave 172.25.254.201:6379
172.25.254.201 6379 @ mymaster 172.25.254.100 6379
39319:X 05 Aug 2024 20:53:16.812 * Sentinel new configuration saved on disk
39319:X 05 Aug 2024 20:53:16.812 * +slave slave 172.25.254.200:6379
172.25.254.200 6379 @ mymaster 172.25.254.100 6379
39319:X 05 Aug 2024 20:53:16.813 * Sentinel new configuration saved on disk
39319:X 05 Aug 2024 20:53:41.769 * +sentinel sentinel
4fe1dcbe25a801e75d6edfc5b0a8517bfa7992c3 172.25.254.200 26379 @ mymaster
172.25.254.100 6379
39319:X 05 Aug 2024 20:53:41.771 * Sentinel new configuration saved on disk
39319:X 05 Aug 2024 20:53:57.227 * +sentinel sentinel
83f928aafe317a5f9081eea8fc5c383ff31c55ef 172.25.254.201 26379 @ mymaster
172.25.254.100 6379
39319:X 05 Aug 2024 20:53:57.229 * Sentinel new configuration saved on disk
...
####在关闭node1的同时会更新以下信息。
1315:X 26 Aug 2024 21:03:51.142 # +sdown slave 172.25.254.10:6379 172.25.254.10 6379 @ mymaster 172.25.254.20 6379
1315:X 26 Aug 2024 21:04:59.710 # +sdown sentinel 3c6130e005b5f06cfe33f7f8f7017e603397d264 172.25.254.20 26379 @ mymaster 172.25.254.20 6379
####在开启node1的同时会更新以下信息。
1315:X 26 Aug 2024 21:06:51.772 # -sdown slave 172.25.254.10:6379 172.25.254.10 6379 @ mymaster 172.25.254.20 6379
1315:X 26 Aug 2024 21:07:01.797 * +convert-to-slave slave 172.25.254.10:6379 172.25.254.10 6379 @ mymaster 172.25.254.20 6379

测试：
在开一个master节点终端
[root@node1 ~]# redis-cli
127.0.0.1:6379> SHUTDOWN
[root@node2 ~]# redis-cli 
127.0.0.1:6379> info replication
# Replication
role:master
connected_slaves:1
slave0:ip=172.25.254.30,port=6379,state=online,offset=60707,lag=1
master_failover_state:no-failover
master_replid:b67a47faa211ce0c3959ea315015da679c5db936
master_replid2:23035f194cb43a56b8a24e0c92fc8b8ce17c1f1d
master_repl_offset:60707
second_repl_offset:34968
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:15
repl_backlog_histlen:60693

#开启node1检验效果
[root@node1 ~]# /etc/init.d/redis_6379 start
Starting Redis server...
[root@node2 ~]# redis-cli 
127.0.0.1:6379> info replication
# Replication
role:master
connected_slaves:2
slave0:ip=172.25.254.30,port=6379,state=online,offset=69697,lag=0
slave1:ip=172.25.254.10,port=6379,state=online,offset=69697,lag=1
master_failover_state:no-failover
master_replid:b67a47faa211ce0c3959ea315015da679c5db936
master_replid2:23035f194cb43a56b8a24e0c92fc8b8ce17c1f1d
master_repl_offset:69697
second_repl_offset:34968
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:15
repl_backlog_histlen:69683

6.4. 在整个架构中可能会出现的问题

问题：在生产环境中如果master和slave中的网络出现故障，由于哨兵的存在会把master提出去当网络恢复后，master发现环境发生改变，master就会把自己的身份转换成slave;master变成slave后会把网络故障那段时间写入自己中的数据清掉，这样数据就丢失了。解决： master在被写入数据时会持续连接slave，mater确保有2个slave可以写入我才允许写入;如果slave数量少于2个便拒绝写入.

#在matster中设定
[root@node2 ~]# redis-cli 
127.0.0.1:6379> CONFIG GET min-slaves-to-write
1) "min-slaves-to-write"
2) "0"
127.0.0.1:6379> CONFIG set min-slaves-to-write 2
OK
127.0.0.1:6379> CONFIG GET min-slaves-to-write
1) "min-slaves-to-write"
2) "2"

#如果要永久保存写到配置文件中
[root@node2 ~]# /etc/redis/6379.conf
....
# even if no authentication is configured.
protected-mode no
min-slaves-to-write 2   #添加这行。

七. Redis Cluster（无中心化设计）

7.1. Redis Cluster 工作原理

在哨兵sentinel机制中，可以解决redis高可用问题，即当master故障后可以自动将slave提升为master，从而可以保证redis服务的正常使用，但是无法解决redis单机写入的瓶颈问题，即单机redis写入性能受;限于单机的内存大小、并发数量、网卡速率等因素。 redis 3.0版本之后推出了无中心架构的redis cluster机制，在无中心的redis集群当中，其每个节点保存当前节点数据和整个集群状态,每个节点都和其他所有节点连接.

7.2. Redis Cluster特点如下

所有Redis节点使用(PING机制)互联;
集群中某个节点的是否失效，是由整个集群中超过半数的节点监测都失效，才能算真正的失效;
客户端不需要proxy即可直接连接redis，应用程序中需要配置有全部的redis服务器IP;
redis cluster把所有的redis node 平均映射到 0-16383个槽位(slot)上，读写需要到指定的redis node上进行操作，因此有多少个redis node相当于redis 并发扩展了多少倍，每个redis node 承担16384/N个槽位;
Redis cluster预先分配16384个(slot)槽位，当需要在redis集群中写入一个key -value的时候，会使用CRC16(key) mod 16384之后的值，决定将key写入值哪一个槽位从而决定写入哪一个Redis节点上，从而有效解决单机瓶颈。

7.3. Redis cluster 主从架构

Redis cluster的架构虽然解决了并发的问题，但是又引入了一个新的问题，每个Redis master的高可用如何解决？那就是对每个master 节点都实现主从复制,从而实现 redis 高可用性

7.4. 创建redis cluster的前提

1.每个redis node节点采用相同的硬件配置、相同的密码、相同的redis版本。 2.每个节点必须开启的参数

cluster-enabled yes #必须开启集群状态，开启后redis进程会有cluster显示
cluster-config-file nodes-6380.conf #此文件有redis cluster集群自动创建和维护，不需要任何手动操作

3.所有redis服务器必须没有任何数据 4.先启动为单机redis且没有任何key value

7.5. 部署redis cluster

查看端口图片：

#环境准备：之前用的源码安装，现在试试rpm安装
#准备六台机子
node1-172.25.254.10
node2-172.25.254.20
node3-172.25.254.30
node4-172.25.254.110
node5-172.25.254.120
node6-172.25.254.130

#连接xshell使用键盘发送所有会话功能，可以同一时间输入相同的指令到机子里面
#在所有redis主机中
[root@node1 ~]# yum install redis -y
[root@redis-masterx ~]# vim /etc/redis/redis.conf
....
bind * -::*  #打开端口
masterauth "123456" #集群主从认证
requirepass "123456" #redis登陆密码 redis-cli 命令连接redis后要
用“auth 密码”进行认证
cluster-enabled yes #开启cluster集群功能
cluster-config-file nodes-6379.conf #指定集群配置文件
cluster-node-timeout 15000 #节点加入集群的超时时间单位是ms
[root@redis-master1 ~]# systemctl enable --now redis.service
[root@node1 ~]# netstat -antlupe | grep redis
[root@redis-master1 ~]# redis-cli
127.0.0.1:6379> auth 123456
OK
127.0.0.1:6379> info
# Cluster
cluster_enabled:1

创建redis-cluster
[root@redis-master1 ~]# redis-cli --cluster create -a 123456 \
> 172.25.254.10:6379 172.25.254.20:6379 172.25.254.30:6379 \
> 172.25.254.110:6379 172.25.254.120:6379 172.25.254.130:6379 \
> --cluster-replicas 1
>>> Performing hash slots allocation on 6 nodes...
Master[0] -> Slots 0 - 5460 #哈希槽分配
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
Adding replica 172.25.254.120:6379 to 172.25.254.10:6379 #主从分配情况
Adding replica 172.25.254.130:6379 to 172.25.254.20:6379
Adding replica 172.25.254.110:6379 to 172.25.254.30:6379
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[0-5460] (5461 slots) master
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[5461-10922] (5462 slots) master
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[10923-16383] (5461 slots) master
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
replicates ba504e78f14df5944280f9035543277a0cf5976b
Can I set the above configuration? (type 'yes' to accept): yes
>>> Performing Cluster Check (using node 172.25.254.10:6379)
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
[OK] All nodes agree about slots configuration.
>>> Check for open slots... #检查打开的哈希槽位
>>> Check slots coverage... #检查槽位覆盖范围
[OK] All 16384 slots covered. #所有槽位分配完成
#配置文件位置
[root@redis-master1 ~]# ll /var/lib/redis/nodes-6379.conf


检测redis集群状态
[root@redis-master1 ~]# redis-cli -a 123456 --cluster info 172.25.254.10:6379
#查看集群状态
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
172.25.254.10:6379 (5ab2e93f...) -> 0 keys | 5461 slots | 1 slaves.
172.25.254.20:6379 (ba504e78...) -> 0 keys | 5462 slots | 1 slaves.
172.25.254.30:6379 (1fcaeb1d...) -> 0 keys | 5461 slots | 1 slaves.
[OK] 0 keys in 3 masters.
0.00 keys per slot on average.
[root@redis-master1 ~]# redis-cli -a 123456 cluster info
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
cluster_state:ok
cluster_slots_assigned:16384
cluster_slots_ok:16384
cluster_slots_pfail:0
cluster_slots_fail:0
cluster_known_nodes:6
cluster_size:3
cluster_current_epoch:6
cluster_my_epoch:1
cluster_stats_messages_ping_sent:821
cluster_stats_messages_pong_sent:801
cluster_stats_messages_sent:1622
cluster_stats_messages_ping_received:796
cluster_stats_messages_pong_received:821
cluster_stats_messages_meet_received:5
cluster_stats_messages_received:1622
[root@redis-master1 ~]# redis-cli -a 123456 --cluster check 172.25.254.10:6379
#检测集群
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
172.25.254.10:6379 (5ab2e93f...) -> 0 keys | 5461 slots | 1 slaves.
172.25.254.20:6379 (ba504e78...) -> 0 keys | 5462 slots | 1 slaves.
172.25.254.30:6379 (1fcaeb1d...) -> 0 keys | 5461 slots | 1 slaves.
[OK] 0 keys in 3 masters.
0.00 keys per slot on average.
>>> Performing Cluster Check (using node 172.25.254.10:6379)
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered

写入数据
[root@redis-master1 ~]# redis-cli -a 123456
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
127.0.0.1:6379> set key1 value1   #被分配到20的hash槽位上
(error) MOVED 9189 172.25.254.20:6379
[root@redis-master2 ~]# redis-cli -a 123456
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
127.0.0.1:6379> set name 赛罗
OK

7.6. 集群扩容

#环境配置
#添加两台机子
node7-172.25.254.50
node8-172.25.254.150

#下载软件
dnf install redis -y

#node1传送配置文件
[root@node1 ~]# scp /etc/redis/redis.conf root@172.25.254.50:/etc/redis/redis.conf
The authenticity of host '172.25.254.50 (172.25.254.50)' can't be established.
ED25519 key fingerprint is SHA256:oBPEX0nYpWQYjJ8OGpbvh+YBXeynOKk0hh0gq7trBAA.
This host key is known by the following other names/addresses:
    ~/.ssh/known_hosts:1: 172.25.254.20
    ~/.ssh/known_hosts:4: 172.25.254.30
Are you sure you want to continue connecting (yes/no/[fingerprint])? yes
Warning: Permanently added '172.25.254.50' (ED25519) to the list of known hosts.
root@172.25.254.50's password: 
redis.conf                                  100%   92KB   3.6MB/s   00:00    
[root@node1 ~]# scp /etc/redis/redis.conf root@172.25.254.150:/etc/redis/redis.conf
The authenticity of host '172.25.254.150 (172.25.254.150)' can't be established.
ED25519 key fingerprint is SHA256:oBPEX0nYpWQYjJ8OGpbvh+YBXeynOKk0hh0gq7trBAA.
This host key is known by the following other names/addresses:
    ~/.ssh/known_hosts:1: 172.25.254.20
    ~/.ssh/known_hosts:4: 172.25.254.30
    ~/.ssh/known_hosts:5: 172.25.254.50
Are you sure you want to continue connecting (yes/no/[fingerprint])? yes
Warning: Permanently added '172.25.254.150' (ED25519) to the list of known hosts.
root@172.25.254.150's password: 
redis.conf                                  100%   92KB   6.6MB/s   00:00   

#机子启动
[root@node7 ~]# systemctl enable --now redis.service

#添加master
[root@redis-master1 ~]# redis-cli -a 123456 --cluster add-node 172.25.254.50:6379 172.25.254.10:6379
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
>>> Adding node 172.25.254.40:6379 to cluster 172.25.254.10:6379
>>> Performing Cluster Check (using node 172.25.254.10:6379)
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
>>> Getting functions from cluster
>>> Failed retrieving Functions from the cluster, skip this step as Redis version
do not support function command (error = 'ERR unknown command `FUNCTION`, with
args beginning with: `DUMP`, ')

>>> Send CLUSTER MEET to node 172.25.254.40:6379 to make it join the cluster.
[OK] New node added correctly.
[root@redis-master1 ~]# redis-cli -a 123456 --cluster info 172.25.254.10:6379
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
172.25.254.10:6379 (5ab2e93f...) -> 0 keys | 5461 slots | 1 slaves.
172.25.254.40:6379 (009571cb...) -> 0 keys | 0 slots | 0 slaves.
172.25.254.20:6379 (ba504e78...) -> 1 keys | 5462 slots | 1 slaves.
172.25.254.30:6379 (1fcaeb1d...) -> 0 keys | 5461 slots | 1 slaves.
[OK] 1 keys in 4 masters.
0.00 keys per slot on average.

#分配槽位
[root@redis-master1 ~]# redis-cli -a 123456 --cluster reshard 172.25.254.10:6379
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
>>> Performing Cluster Check (using node 172.25.254.10:6379)
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
M: 009571cb206a89afa6658b60b2d403136056ac09 172.25.254.40:6379
slots: (0 slots) master
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
How many slots do you want to move (from 1 to 16384)? 4096  #一共16384分个四个，一共4096.
What is the receiving node ID? 009571cb206a89afa6658b60b2d403136056ac09  #给50自己的ip
Please enter all the source node IDs.
Type 'all' to use all the nodes as source nodes for the hash slots.
Type 'done' once you entered all the source nodes IDs.
Source node #1: all  #所有人都给他分槽位

#添加salve
[root@redis-master1 ~]# redis-cli -a 123456 --cluster add-node
172.25.254.150:6379 172.25.254.10:6379 --cluster-slave --cluster-master-id 8eb1a727f1b2d9d1622f95605a64b2f7bf2a5b4a
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
>>> Adding node 172.25.254.140:6379 to cluster 172.25.254.10:6379
>>> Performing Cluster Check (using node 172.25.254.10:6379)
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[1365-5460] (4096 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
M: 009571cb206a89afa6658b60b2d403136056ac09 172.25.254.40:6379
slots:[0-1364],[5461-6826],[10923-12287] (4096 slots) master
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[6827-10922] (4096 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[12288-16383] (4096 slots) master
1 additional replica(s)
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
>>> Send CLUSTER MEET to node 172.25.254.140:6379 to make it join the cluster.
Waiting for the cluster to join
>>> Configure node as replica of 172.25.254.40:6379.
[OK] New node added correctly.

7.7. clsuter集群维护

添加节点的时候是先添加node节点到集群，然后分配槽位，删除节点的操作与添加节点的操作正好相反，是先将被删除的Redis node上的槽位迁移到集群中的其他Redis node节点上，然后再将其删除，如果一个Redis node节点上的槽位没有被完全迁移，删除该node的时候会提示有数据且无法删除。

#删除slave
[root@node1 ~]# redis-cli -a 123456 --cluster del-node 172.25.254.150:6379 67923656726e54c2cb375b8b6b69a3f081f1e808  #加自己的id
[root@node1 ~]# redis-cli -a 123456 --cluster check 172.25.254.10:6379

#移除要下线主机的哈希槽位
[root@redis-master2 ~]# redis-cli -a 123456 --cluster reshard 172.25.254.10:6379
Warning: Using a password with '-a' or '-u' option on the command line interface
may not be safe.
>>> Performing Cluster Check (using node 172.25.254.20:6379)
M: ba504e78f14df5944280f9035543277a0cf5976b 172.25.254.20:6379
slots:[6827-10922] (4096 slots) master
1 additional replica(s)
M: 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7 172.25.254.30:6379
slots:[12288-16383] (4096 slots) master
1 additional replica(s)
M: 009571cb206a89afa6658b60b2d403136056ac09 172.25.254.40:6379
slots:[0-1364],[5461-6826],[10923-12287] (4096 slots) master
1 additional replica(s)
S: c20c9b5465b2e64868161c0e285d55bc81358ba4 172.25.254.110:6379
slots: (0 slots) slave
replicates 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7
S: d458f34fa900d83212c021dc1e65396e490b5495 172.25.254.120:6379
slots: (0 slots) slave
replicates 5ab2e93f4f0783983676f7bd118efaacfb202bd1
M: 5ab2e93f4f0783983676f7bd118efaacfb202bd1 172.25.254.10:6379
slots:[1365-5460] (4096 slots) master
1 additional replica(s)
S: 83d7a82fe896cf9f4d8212cb533058659bba16ce 172.25.254.130:6379
slots: (0 slots) slave
replicates ba504e78f14df5944280f9035543277a0cf5976b
S: 86a4a8fb08e70e41b5a30f829deb983d23854ea7 172.25.254.140:6379
slots: (0 slots) slave
replicates 009571cb206a89afa6658b60b2d403136056ac09
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
How many slots do you want to move (from 1 to 16384)?   4096
What is the receiving node ID? 1fcaeb1dd936b46f4ea1efe4330c54195e66acf7  #把空间给谁，就是谁的id
Please enter all the source node IDs.
Type 'all' to use all the nodes as source nodes for the hash slots.
Type 'done' once you entered all the source nodes IDs.
Source node #1: 5ab2e93f4f0783983676f7bd118efaacfb202bd1
Source node #2: done

#删除master
[root@node1 ~]# redis-cli -a 123456 --cluster del-node 172.25.254.50:6379 c890232037fc544aed8062fab60f6683489b5c3e
Warning: Using a password with '-a' or '-u' option on the command line interface may not be safe.
>>> Removing node c890232037fc544aed8062fab60f6683489b5c3e from cluster 172.25.254.50:6379
>>> Sending CLUSTER FORGET messages to the cluster...
>>> Sending CLUSTER RESET SOFT to the deleted node.
[root@node8 ~]# redis-cli -a 123456 --cluster check 172.25.254.10:6379