NoSQL之 Redis集群

案例概述

单节点Redis服务器带来的问题

• 单点故障，服务不可用
• 无法处理大量的并发数据请求
• 数据丢失—大灾难

解决方法

• 搭建Redis集群

Redis集群知识点

Redis集群介绍

• Redis集群是一个提供在多个Redis间节点间共享数据的程序集
• Redis集群并不支持处理多个keys的命令,因为这需要在不同的节点间移动数据,从而达不到像Redis那样的性能,在高负载的情况下可能会导致不可预料的错误
• Redis集群通过分区来提供—定程度的可用性,在实际环境中当某个节点宕机或者不可达的情况下可继续处理命令

Redis集群的优势

• 自动分割数据到不同的节点上
• 整个集群的部分节点失败或者不可达的情况下能够继续处理命令

Redis集群的实现方法

• 有客户端分片
• 代理分片
• 服务器端分片

Redis-Cluster数据分片

1：Redis集群没有使用一致性hash,而是引入了哈希槽概念

2：Redis 集群有16384个哈希槽

3：每个key通过CRC16校验后对16384取模来决定放置槽

4：集群的每个节点负责一部分哈希槽

5：以3个节点组成的集群为例

• 节点A包含0到5500号哈希槽
• 节点B包含5501到11000号哈希槽
• 节点C包含11001到16383号哈希槽支持添加或者删除节点

6：添加删除节点无需停止服务

• 如果想新添加个节点D,需要移动节点A,B,C中的部分槽到D上
• 如果想移除节点A，需要将A中的槽移到B和C节点上，再将没有任何槽的A节点从集群中移除

Redis-Cluster的主从复制模型

• 集群中具有A,B，C三个节点,如果节点B失败了，整个集群就会因缺少5501-11000这个范围的槽而不可用
• 为每个节点添加一个从节点A1，B1，C1,整个集群便有三个master节点和三个slave节点组成在节点B失败后，集群便会选举B1为新的主节点继续服务
• 当B和B1都失败后，集群将不可用

案例拓扑图

具体步骤

1：每台服务器导入redis软件包，其中一台主redis要有群集文件

2：解压缩redis软件包，6台服务器同步操作

[root@redis1 ~]# tar zxvf redis-5.0.4.tar.gz

3：配置安装，6台服务器同步操作

[root@redis1 ~]# cd redis-5.0.4/
[root@redis1 redis-5.0.4]# make
[root@redis1 redis-5.0.4]# make PREFIX=/usr/local/redis install

4：链接快捷命令，6台服务器同步操作

[root@redis1 ~]# ln -s /usr/local/redis/bin/* /usr/local/bin/

5：设置Redis相关配置文件，6台服务器同步操作

[root@redis1 ~]# cd redis-5.0.4/utils/
[root@redis1 utils]# ./install_server.sh 
[root@redis1 utils]# netstat -anptu | grep redis
tcp        0      0 127.0.0.1:6379        0.0.0.0:*   LISTEN    59531/redis-server 

6：修改主配置文件，6台服务器同步操作

[root@slave3 ~]# vi /etc/redis/6379.conf 
cluster-enabled yes
appendonly yes
cluster-config-file nodes-6379.conf
cluster-node-timeout 15000
cluster-require-full-coverage yes

修改IP地址需要每台服务器单独修改

[root@redis1 ~]# vi /etc/redis/6379.conf 
bind 20.0.0.10

[root@redis2 ~]# vi /etc/redis/6379.conf
bind 20.0.0.11

[root@redis3 ~]# vi /etc/redis/6379.conf
bind 20.0.0.12

[root@slave1 ~]# vi /etc/redis/6379.conf 
bind 20.0.0.13

[root@slave2 ~]# vi /etc/redis/6379.conf 
bind 20.0.0.14

[root@slave3 ~]# vi /etc/redis/6379.conf 
bind 20.0.0.15

7：重启服务，6台服务器同步操作

[root@redis1 ~]# /etc/init.d/redis_6379 restart
Stopping ...
Waiting for Redis to shutdown ...
Redis stopped
Starting Redis server...
[root@redis1 ~]# netstat -anptu | grep redis

8：在20.0.0.10服务器上，使用脚本创建群集

[root@redis1 ~]# yum -y install ruby rubygems
[root@redis1 ~]# gem install redis-3.2.0.gem

9：建立群集

[root@redis1 ~]# redis-cli --cluster create --cluster-replicas 1 20.0.0.10:6379 20.0.0.11:6379 20.0.0.12:6379 20.0.0.13:6379 20.0.0.14:6379 20.0.0.15:6379

>>> Performing hash slots allocation on 6 nodes...
Master[0] -> Slots 0 - 5460
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
Adding replica 20.0.0.14:6379 to 20.0.0.10:6379
Adding replica 20.0.0.15:6379 to 20.0.0.11:6379
Adding replica 20.0.0.13:6379 to 20.0.0.12:6379
M: 2ce100a99c214bfac23ae1c31b8068e93506615b 20.0.0.10:6379
   slots:[0-5460] (5461 slots) master
M: 5f5ff79b9301f10e175aea6604ff80372d6c97a8 20.0.0.11:6379
   slots:[5461-10922] (5462 slots) master
M: a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 20.0.0.12:6379
   slots:[10923-16383] (5461 slots) master
S: b378987085da43974359cb02bcbc9443e808bb13 20.0.0.13:6379
   replicates a6234f074de2ee05b4f7c1ba96c8a3e084104dd4
S: 5d27b053b2607dfed82f79d29aa2e5c3b8846bc2 20.0.0.14:6379
   replicates 2ce100a99c214bfac23ae1c31b8068e93506615b
S: 1d2b8d8fd38f5e7c2dce00ee43e13f162a18508d 20.0.0.15:6379
   replicates 5f5ff79b9301f10e175aea6604ff80372d6c97a8
Can I set the above configuration? (type 'yes' to accept): yes
>>> Nodes configuration updated
>>> Assign a different config epoch to each node
>>> Sending CLUSTER MEET messages to join the cluster
Waiting for the cluster to join
..
>>> Performing Cluster Check (using node 20.0.0.10:6379)
M: 2ce100a99c214bfac23ae1c31b8068e93506615b 20.0.0.10:6379
   slots:[0-5460] (5461 slots) master
   1 additional replica(s)
M: a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 20.0.0.12:6379
   slots:[10923-16383] (5461 slots) master
   1 additional replica(s)
S: 1d2b8d8fd38f5e7c2dce00ee43e13f162a18508d 20.0.0.15:6379
   slots: (0 slots) slave
   replicates 5f5ff79b9301f10e175aea6604ff80372d6c97a8
S: b378987085da43974359cb02bcbc9443e808bb13 20.0.0.13:6379
   slots: (0 slots) slave
   replicates a6234f074de2ee05b4f7c1ba96c8a3e084104dd4
M: 5f5ff79b9301f10e175aea6604ff80372d6c97a8 20.0.0.11:6379
   slots:[5461-10922] (5462 slots) master
   1 additional replica(s)
S: 5d27b053b2607dfed82f79d29aa2e5c3b8846bc2 20.0.0.14:6379
   slots: (0 slots) slave
   replicates 2ce100a99c214bfac23ae1c31b8068e93506615b
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

10：测试群集

登录写入一个数据

[root@redis1 ~]# redis-cli -h 20.0.0.10 -p 6379 -c
20.0.0.10:6379> set aaa 111
-> Redirected to slot [10439] located at 20.0.0.11:6379
OK

另一台主服务器获取数据

[root@redis3 ~]# redis-cli -h 20.0.0.13 -p 6379 -c
20.0.0.13:6379> get aaa
-> Redirected to slot [10439] located at 20.0.0.11:6379
"111"

登录不同服务器获取数据

[root@redis2 ~]# redis-cli -h 20.0.0.13 -p 6379 -c
20.0.0.13:6379> get aaa
-> Redirected to slot [10439] located at 20.0.0.11:6379
"111"

查看集群状态

20.0.0.11:6379> cluster info
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:2
cluster_stats_messages_ping_sent:5018
cluster_stats_messages_pong_sent:4927
cluster_stats_messages_meet_sent:2
cluster_stats_messages_sent:9947
cluster_stats_messages_ping_received:4923
cluster_stats_messages_pong_received:5020
cluster_stats_messages_meet_received:4
cluster_stats_messages_received:9947

查看节点状态

20.0.0.11:6379> cluster nodes
5d27b053b2607dfed82f79d29aa2e5c3b8846bc2 20.0.0.14:6379@16379 slave 2ce100a99c214bfac23ae1c31b8068e93506615b 0 1608200801776 1 connected
b378987085da43974359cb02bcbc9443e808bb13 20.0.0.13:6379@16379 slave a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 0 1608200799000 3 connected
a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 20.0.0.12:6379@16379 master - 0 1608200801000 3 connected 10923-16383
1d2b8d8fd38f5e7c2dce00ee43e13f162a18508d 20.0.0.15:6379@16379 slave 5f5ff79b9301f10e175aea6604ff80372d6c97a8 0 1608200802783 6 connected
5f5ff79b9301f10e175aea6604ff80372d6c97a8 20.0.0.11:6379@16379 myself,master - 0 1608200801000 2 connected 5461-10922
2ce100a99c214bfac23ae1c31b8068e93506615b 20.0.0.10:6379@16379 master - 0 1608200801000 1 connected 0-5460

11：宕机测试

主服务器停止服务

[root@redis1 ~]# pkill redis

查看状态

20.0.0.11:6379> cluster nodes
5d27b053b2607dfed82f79d29aa2e5c3b8846bc2 20.0.0.14:6379@16379 slave 2ce100a99c214bfac23ae1c31b8068e93506615b 0 1608201296530 1 connected
b378987085da43974359cb02bcbc9443e808bb13 20.0.0.13:6379@16379 slave a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 0 1608201297542 3 connected
a6234f074de2ee05b4f7c1ba96c8a3e084104dd4 20.0.0.12:6379@16379 master - 0 1608201298547 3 connected 10923-16383
1d2b8d8fd38f5e7c2dce00ee43e13f162a18508d 20.0.0.15:6379@16379 slave 5f5ff79b9301f10e175aea6604ff80372d6c97a8 0 1608201299558 6 connected
5f5ff79b9301f10e175aea6604ff80372d6c97a8 20.0.0.11:6379@16379 myself,master - 0 1608201294000 2 connected 5461-10922
2ce100a99c214bfac23ae1c31b8068e93506615b 20.0.0.10:6379@16379 master,fail - 1608201280782 1608201277358 1 disconnected 0-5460
备份20.0.0.15以上线