Redis集群搭建与简单使用

目录

Redis集群搭建与简单使用

一、安装过程

1. 下载并解压

2. 编译安装

3. 创建 Redis 节点

4. 启动各个节点

5. 检查 redis 启动情况

6.设置防火墙&＃xff0c;开放集群端口

7.创建集群

二、集群测试

1.存取测试

2.宕机测试

三、节点添加

1.新增主节点

2.新增从节点

四、节点移除

1.移除主节点

2.移除从节点

---

 

 

介绍安装环境与版本

用两台虚拟机模拟6个节点&＃xff0c;一台机器3个节点&＃xff0c;创建出3 master、3 salve 环境&＃xff08;必须要3个或以上的主节点&＃xff0c;否则在创建集群时会失败&＃xff09;。

redis 采用 redis-5.0.3 版本&＃xff08;5.0版本不需要安装Ruby&＃xff0c;使用较为方便&＃xff0c;推荐安装&＃xff09;。

两台虚拟机都是 CentOS &＃xff0c;一台 CentOS7.5 &＃xff08;IP:192.168.92.100&＃xff09;&＃xff0c;一台 CentOS7.5&＃xff08;IP:192.168.92.101&＃xff09; 。

一、安装过程

1. 下载并解压

[root&＃64;localhost software]# cd /root/software
[root&＃64;localhost software]# wget http://download.redis.io/releases/redis-5.0.3.tar.gz
[root&＃64;localhost software]# tar -zxvf redis-5.0.3.tar.gz

2. 编译安装

[root&＃64;localhost software]# cd redis-5.0.3
[root&＃64;localhost redis-5.0.3]# make && make install

3. 创建 Redis 节点

首先在 192.168.92.100 机器上 /root/software/redis-5.0.3 目录下创建 redis_cluster 目录&＃xff1b;

[root&＃64;localhost redis-5.0.3]# mkdir redis_cluster

在 redis_cluster 目录下&＃xff0c;创建名为7000、7001、7002的目录&＃xff0c;并将 redis.conf 拷贝到这三个目录中

[root&＃64;localhost redis_cluster]# mkdir 7000 7001 7002
[root&＃64;localhost redis-5.0.3]# cp redis.conf redis_cluster/7000
[root&＃64;localhost redis-5.0.3]# cp redis.conf redis_cluster/7001
[root&＃64;localhost redis-5.0.3]# cp redis.conf redis_cluster/7002　　

分别修改这三个配置文件&＃xff0c;修改如下内容

port 7000 //端口7000,7001,7002
bind 本机ip //默认ip为127.0.0.1 需要改为其他节点机器可访问的ip 否则创建集群时无法访问对应的端口&＃xff0c;无法创建集群
daemonize yes //redis后台运行
pidfile /var/run/redis_7000.pid //pidfile文件对应7000,7001,7002
cluster-enabled yes //开启集群 把注释#去掉
cluster-config-file nodes_7000.conf //集群的配置 配置文件首次启动自动生成 7000,7001,7002
cluster-node-timeout 15000 //请求超时 默认15秒&＃xff0c;可自行设置
appendonly yes //aof日志开启 有需要就开启&＃xff0c;它会每次写操作都记录一条日志

• 接着在另外一台机器上&＃xff08;192.168.92.101&＃xff09;重复以上三步&＃xff08;1. 下载并解压&＃xff1b;2. 编译安装&＃xff1b;3. 创建 Redis 节点&＃xff09;的操作&＃xff0c;只是把目录改为7003、7004、7005&＃xff0c;对应的配置文件也按照这个规则修改即可

4. 启动各个节点

第一台机器&＃xff08;192.168.92.100&＃xff09;上执行


[root&＃64;localhost 7000]# redis-server redis.conf
[root&＃64;localhost 7001]# redis-server redis.conf
[root&＃64;localhost 7002]# redis-server redis.conf

另外一台机器&＃xff08;192.168.92.101&＃xff09;上执行


[root&＃64;localhost 7003]# redis-server redis.conf
[root&＃64;localhost 7004]# redis-server redis.conf
[root&＃64;localhost 7005]# redis-server redis.conf

注&＃xff1a;可以写一个命令脚本start-all.sh &＃xff0c;来启动各个节点

cd 7000
redis-server redis.conf
cd ..
cd 7001
redis-server redis.conf
cd ..
cd 7002
redis-server redis.conf

设置权限启动

[root&＃64;localhost redis-cluster]# chmod 777 start-all.sh 
[root&＃64;localhost redis-cluster]# ./start-all.sh

5. 检查 redis 启动情况

##一台机器&＃xff08;192.168.92.100&＃xff09;

[root&＃64;localhost redis-5.0.3]# ps -ef | grep redis
root 61020 1 0 02:14 ? 00:00:01 redis-server 127.0.0.1:7000 [cluster]
root 61024 1 0 02:14 ? 00:00:01 redis-server 127.0.0.1:7001 [cluster]
root 61029 1 0 02:14 ? 00:00:01 redis-server 127.0.0.1:7002 [cluster] [root&＃64;localhost redis-5.0.3]# netstat -tnlp | grep redis
tcp 0 0 127.0.0.1:17000 0.0.0.0:* LISTEN 61020/redis-server
tcp 0 0 127.0.0.1:17001 0.0.0.0:* LISTEN 61024/redis-server
tcp 0 0 127.0.0.1:17002 0.0.0.0:* LISTEN 61029/redis-server
tcp 0 0 127.0.0.1:7000 0.0.0.0:* LISTEN 61020/redis-server
tcp 0 0 127.0.0.1:7001 0.0.0.0:* LISTEN 61024/redis-server
tcp 0 0 127.0.0.1:7002 0.0.0.0:* LISTEN 61029/redis-server

##另外一台机器 

[root&＃64;localhost redis-5.0.3]# ps -ef | grep redis
root 9957 1 0 02:32 ? 00:00:01 redis-server 127.0.0.1:7003 [cluster]
root 9964 1 0 02:32 ? 00:00:01 redis-server 127.0.0.1:7004 [cluster]
root 9971 1 0 02:32 ? 00:00:01 redis-server 127.0.0.1:7005 [cluster]
root 10065 4744 0 02:38 pts/0 00:00:00 grep --color&＃61;auto redis
[root&＃64;localhost redis-5.0.3]# netstat -tlnp | grep redis
tcp 0 0 127.0.0.1:17003 0.0.0.0:* LISTEN 9957/redis-server 1
tcp 0 0 127.0.0.1:17004 0.0.0.0:* LISTEN 9964/redis-server 1
tcp 0 0 127.0.0.1:17005 0.0.0.0:* LISTEN 9971/redis-server 1
tcp 0 0 127.0.0.1:7003 0.0.0.0:* LISTEN 9957/redis-server 1
tcp 0 0 127.0.0.1:7004 0.0.0.0:* LISTEN 9964/redis-server 1
tcp 0 0 127.0.0.1:7005 0.0.0.0:* LISTEN 9971/redis-server 1

6.设置防火墙&＃xff0c;开放集群端口

两台机器的防火墙我们直接关掉&＃xff08;如果不关掉&＃xff0c;在下面创建集群的时候&＃xff0c;会提示&＃xff1a;Could not connect to Redis at 192.168.92.101:7003: No route to host&＃xff09;

[root&＃64;localhost redis-5.0.3]# systemctl stop firewalld.service

7.创建集群

Redis 官方提供了 redis-cli &＃xff08;5.0之前的版本是使用的 redis-trib.rb &＃xff0c;需要安装ruby&＃xff0c;不推荐&＃xff09;这个工具&＃xff0c;可以直接在命令行中使用了。使用下面这个命令即可完成安装。

[root&＃64;localhost redis-5.0.3]# redis-cli --cluster create 192.168.92.100:7000 192.168.92.100:7001 192.168.92.100:7002 192.168.92.101:7003 192.168.92.101:7004 192.168.92.101:7005 --cluster-replicas 1

• redis-cli --cluster create&＃xff0c;创建一个新的集群&＃xff1b;
• 前三个 ip:port 为第一台机器的节点&＃xff0c;剩下三个为第二台机器&＃xff1b;
• 选项 --cluster-replicas 1 表示为集群中的每个主节点创建一个从节点。

接着&＃xff0c; redis-cli会输出一份配置信息&＃xff0c;如下&＃xff1a;

>>> Performing hash slots allocation on 6 nodes...
Master[0] -> Slots 0 - 5460
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
Adding replica 192.168.92.101:7004 to 192.168.92.100:7000
Adding replica 192.168.92.100:7002 to 192.168.92.101:7003
Adding replica 192.168.92.101:7005 to 192.168.92.100:7001
M: 5c55309b5a1cd704a056cf80be308964e757f7a1 192.168.92.100:7000slots:[0-5460] (5461 slots) master
M: 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a 192.168.92.100:7001slots:[10923-16383] (5461 slots) master
S: b8b90944c7c65511da0d0b04004d502015f599dc 192.168.92.100:7002replicates c4d9fdc43a459d19bb78378b0f801b70f4dd70ec
M: c4d9fdc43a459d19bb78378b0f801b70f4dd70ec 192.168.92.101:7003slots:[5461-10922] (5462 slots) master
S: 60818181c7a6bb007b3628ef4adf84678359f613 192.168.92.101:7004replicates 5c55309b5a1cd704a056cf80be308964e757f7a1
S: 57b19e309ec580857968f113de4bd2b77aa05ac1 192.168.92.101:7005replicates 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a
Can I set the above configuration? (type &＃39;yes&＃39; to accept):

确认无误后&＃xff0c;输入yes &＃xff0c; redis-cli 会将配置应用到各个节点&＃xff0c;并连接起&＃xff08;join)各个节点&＃xff0c;让各个节点开始通讯。出现如下内容&＃xff0c;说明安装成功。

>>> 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 192.168.92.100:7000)
M: 5c55309b5a1cd704a056cf80be308964e757f7a1 192.168.92.100:7000slots:[0-5460] (5461 slots) master1 additional replica(s)
M: 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a 192.168.92.100:7001slots:[10923-16383] (5461 slots) master1 additional replica(s)
S: 57b19e309ec580857968f113de4bd2b77aa05ac1 192.168.92.101:7005slots: (0 slots) slavereplicates 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a
S: 60818181c7a6bb007b3628ef4adf84678359f613 192.168.92.101:7004slots: (0 slots) slavereplicates 5c55309b5a1cd704a056cf80be308964e757f7a1
S: b8b90944c7c65511da0d0b04004d502015f599dc 192.168.92.100:7002slots: (0 slots) slavereplicates c4d9fdc43a459d19bb78378b0f801b70f4dd70ec
M: c4d9fdc43a459d19bb78378b0f801b70f4dd70ec 192.168.92.101:7003slots:[5461-10922] (5462 slots) master1 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

二、集群测试

1.存取测试

示例&＃xff1a;

访问192.168.92.100的7000端口对应的redis节点

[root&＃64;localhost ~]# redis-cli -c -h 192.168.92.100 -p 7000
192.168.92.100:7000>

参数 -c 可连接到集群&＃xff0c;因为 redis.conf 将 bind 改为了ip地址&＃xff0c;所以 -h 参数不能省略&＃xff0c;-p 参数为端口号。

写入数据&＃xff0c;进行集群的验证

192.168.92.100:7000> set name andy
-> Redirected to slot [5798] located at 192.168.92.101:7003
OK
192.168.92.101:7003>

发现redis set name andy之后重定向到 192.168.92.101机器 redis 7003这个节点

我们在这个集群中任意一个节点去获取该key的值&＃xff0c;假设在192.168.92.100的7002端口对应的redis节点上去获取

[root&＃64;localhost ~]# redis-cli -c -h 192.168.92.100 -p 7002
192.168.92.100:7002> get name
-> Redirected to slot [5798] located at 192.168.92.101:7003
"andy"
192.168.92.101:7003>

 7002节点&＃xff0c;从上面也是自动跳转至7003获取值&＃xff0c;这也是redis cluster的特点&＃xff0c;它是去中心化&＃xff0c;每个节点都是对等的&＃xff0c;连接哪个节点都可以获取和设置数据。

2.宕机测试

现在模拟将7003节点挂掉&＃xff0c;按照redis-cluster原理会选举会将 7003的从节点7002选举为主节点。

[root&＃64;localhost ~]# ps -ef | grep redis
root 13288 1 0 17:01 ? 00:00:17 redis-server 192.168.92.101:7003 [cluster]
root 13293 1 0 17:01 ? 00:00:18 redis-server 192.168.92.101:7004 [cluster]
root 13298 1 0 17:01 ? 00:00:17 redis-server 192.168.92.101:7005 [cluster]
root 13435 13417 0 19:29 pts/1 00:00:00 grep --color&＃61;auto redis
[root&＃64;localhost ~]# kill 13288
[root&＃64;localhost ~]# ps -ef | grep redis
root 13293 1 0 17:01 ? 00:00:18 redis-server 192.168.92.101:7004 [cluster]
root 13298 1 0 17:01 ? 00:00:18 redis-server 192.168.92.101:7005 [cluster]
root 13437 13417 0 19:30 pts/1 00:00:00 grep --color&＃61;auto redis

再查看集群中的7003节点

[root&＃64;localhost ~]# redis-cli --cluster check 192.168.92.101:7003
Could not connect to Redis at 192.168.92.101:7003: Connection refused
[root&＃64;localhost ~]# redis-cli --cluster check 192.168.92.100:7002
Could not connect to Redis at 192.168.92.101:7003: Connection refused
192.168.92.100:7002 (b8b90944...) -> 1 keys | 5462 slots | 0 slaves.
192.168.92.100:7001 (4cbdd0a0...) -> 0 keys | 5461 slots | 1 slaves.
192.168.92.100:7000 (5c55309b...) -> 0 keys | 5461 slots | 1 slaves.
[OK] 1 keys in 3 masters.
0.00 keys per slot on average.
>>> Performing Cluster Check (using node 192.168.92.100:7002)
M: b8b90944c7c65511da0d0b04004d502015f599dc 192.168.92.100:7002slots:[5461-10922] (5462 slots) master
S: 57b19e309ec580857968f113de4bd2b77aa05ac1 192.168.92.101:7005slots: (0 slots) slavereplicates 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a
S: 60818181c7a6bb007b3628ef4adf84678359f613 192.168.92.101:7004slots: (0 slots) slavereplicates 5c55309b5a1cd704a056cf80be308964e757f7a1
M: 4cbdd0a0b1a8e9334ad97a3df577e8ab365e4f7a 192.168.92.100:7001slots:[10923-16383] (5461 slots) master1 additional replica(s)
M: 5c55309b5a1cd704a056cf80be308964e757f7a1 192.168.92.100:7000slots:[0-5460] (5461 slots) master1 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

可以看到集群连接不了7003节点&＃xff0c;而7002由原来的S转换为M节点&＃xff0c;代替了原来的7003节点。

所有会有最后一个说明All 16384 slots covered. 所有哈希槽都可覆盖了&＃xff1b;

我们可以获取name值&＃xff1a;

[root&＃64;localhost ~]# redis-cli -c -h 192.168.92.100 -p 7000
192.168.92.100:7000> get name
-> Redirected to slot [5798] located at 192.168.92.100:7002
"andy"
192.168.92.100:7002>

从7000节点连入&＃xff0c;自动跳转到7002节点&＃xff0c;并且获取name值。

现在我们将7003节点恢复&＃xff0c;看是否会自动加入集群中以及充当的M还是S节点。

[root&＃64;localhost 7003]# redis-server redis.conf
13451:C 15 Feb 2019 19:43:36.259 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
13451:C 15 Feb 2019 19:43:36.260 # Redis version&＃61;5.0.3, bits&＃61;64, commit&＃61;00000000, modified&＃61;0, pid&＃61;13451, just started
13451:C 15 Feb 2019 19:43:36.260 # Configuration loaded

 再check一下7003节点

[root&＃64;localhost redis_cluster]# redis-cli --cluster check 192.168.92.101:7003
192.168.92.100:7000 (7980a27f...) -> 3 keys | 5461 slots | 1 slaves.
192.168.92.100:7001 (255fe370...) -> 0 keys | 5461 slots | 1 slaves.
192.168.92.100:7002 (5ad1bc99...) -> 2 keys | 5462 slots | 1 slaves.
[OK] 5 keys in 3 masters.
0.00 keys per slot on average.
>>> Performing Cluster Check (using node 192.168.92.101:7003)
S: 391fb1127f7ccc30472d3761719272aa1dbe38f1 192.168.92.101:7003slots: (0 slots) slavereplicates 5ad1bc995f4af9c2d8aa862ec3d5902633a29b54
M: 7980a27ff9c1f37d8f185c1419b0445b9da1a32b 192.168.92.100:7000slots:[0-5460] (5461 slots) master1 additional replica(s)
S: 0b05268b52f540dddcfe17f56e17d97eb41b759f 192.168.92.101:7004slots: (0 slots) slavereplicates 7980a27ff9c1f37d8f185c1419b0445b9da1a32b
S: 9adfa4f073c0949861559de79cc19a4fba368307 192.168.92.101:7005slots: (0 slots) slavereplicates 255fe37083527c29fc352486880ecad297d5ff58
M: 255fe37083527c29fc352486880ecad297d5ff58 192.168.92.100:7001slots:[10923-16383] (5461 slots) master1 additional replica(s)
M: 5ad1bc995f4af9c2d8aa862ec3d5902633a29b54 192.168.92.100:7002slots:[5461-10922] (5462 slots) master1 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

可以看到7003节点变成了5ad1bc995f4af9c2d8aa862ec3d5902633a29b54  7002的从节点。

补充说明&＃xff1a;

1.节点的fail

节点的fail是通过集群中超过半数的节点检测失效时才生效&＃xff0c;这里会触发选举&＃xff0c;将slave升级为master

2.集群的fail

• 如果集群任意master挂掉,且当前master没有slave.集群进入fail状态,也可以理解成进群的slot映射[0-16383]不完整时进入fail状态&＃xff08;redis-3.0.0.rc1加入cluster-require-full-coverage参数,默认关闭,打开集群兼容部分失败&＃xff09;&＃xff1b;
• 如果集群超过半数以上master挂掉&＃xff0c;无论是否有slave集群进入fail状态。

三、节点添加

1.新增主节点

待更新...

2.新增从节点

待更新...

四、节点移除

1.移除主节点

待更新...

2.移除从节点

待更新...