redis三主三从集群部署

╰+攻爆jí腚メ 2022-04-11 10:11 306阅读 0赞

## 1.redis集群原理 ##

Redis集群模式主要有2种:

*  主从集群
 *  分布式集群。

Redis为了实现负载均衡，提供集群模式。以三个节点为例，集群模式相当于将1-15000片分片，分为1-5000、5000-10000、10000-15000。每个节点分一段数据片。这样的话，当一个节点宕机后，这个节点没有备份的话，此段分片将不再可以使用。所以，官方推荐，集群内的每个节点都应该配备一个从节点，作为冷备。

官网推荐的模式，是三主三从的集群部署方式。

## 2.redis集群部署 ##

### 2.1解决依赖，获取解压安装包 ###

$ yum -y install gcc #如果没有gcc编译容器则需要安装
     $ wget http://download.redis.io/releases/redis-4.0.11.tar.gz
     $ tar -xf redis-4.0.11.tar.gz

### 2.2 安装，并配置环境变量 ###

#编译，获取可行执行文件
    $ cd redis-4.0.11
    $ make
      
    #在/opt下创建redis目录，将可执行文件，配置文件，日志文件都放在/opt/redis目录下
    $ mkdir  -pv /opt/redis/conf /opt/redis/bin /opt/redis/log  /opt/redis/data/7000   /opt/redis/data/7001
     
    #拷贝常用的二进制文件
    $ cp /opt/software/redis-4.0.11/src/redis-* /opt/redis/bin
     
     
    #删除.c，.o的文件
    $ rm /opt/redis/bin/*.c  /opt/redis/bin/*.o
    
    $ vim /etc/profile  #添加以下内容
    
    export REDIS_HOME=/opt/redis
    export PATH=${REDIS_HOME}/bin:$PATH
    
    #立即生效
    $ source /etc/profile

在第一台机器上配置redis的主，从两个配置文件，下面是主实例的配置文件内容

$ vim /opt/redis/conf/redis7000.conf
    
    bind ${ip1}         #填写第一台机器ip  
    port 7000
    cluster-enabled  yes
    cluster-config-file nodes.conf
    cluster-node-timeout 5000
    appendonly yes
    timeout 3600
    daemonize yes
      
    pidfile /opt/redis/pid/redis7000.pid
    logfile "/opt/redis/log/redis7000.log"
    dir "/opt/redis/data/7000"
      
    rename-command FLUSHALL ""
    rename-command FLUSHDB ""
    rename-command KEYS ""

下面是从实例的配置文件内容，只是端口不一样，和目录位置结构不一样

#这里我们使用简单一点集群配置，从实例配置文件
    $ vim /opt/redis/conf/redis7001.conf
    
    bind ${ip1}
    port 7001
    cluster-enabled  yes
    cluster-config-file nodes.conf
    cluster-node-timeout 5000
    appendonly yes
    timeout 3600
    daemonize yes
      
    pidfile /opt/redis/pid/redis7001.pid
    logfile "/opt/redis/log/redis7001.log"
    dir "/opt/redis/data/7001"
      
    rename-command FLUSHALL ""
    rename-command FLUSHDB ""
    rename-command KEYS ""

### 2.3启动redis实例 ###

配置完后，我们scp到其他两个机器上，并三台机器都要重动，每台启动2个实例，一个主实例，一个从实例。我们以第一台机器为例，进行重启。

#拷贝到其他两台机器上
    $ scp -r /opt/redis ${ip2}:/opt
    $ scp -r /opt/redis ${ip3}:/opt
      
    #启动redis实例,3台机器均要启动
    $ redis-server conf/redis7000.conf
    $ redis-server conf/redis7001.conf

我们可以看到两个进程存在，其他两台机器也要启动，启动方式是一样的。

$ ps -ef |grep redis
    root       21882       1  0 Nov30 ?        00:09:05 redis-server ${ip1}:7000 [cluster]
    root       21891       1  0 Nov30 ?        00:08:45 redis-server ${ip1}:7001 [cluster]

## 3. 创建集群 ##

### 3.1安装ruby ###

创建Redis集群需要借助安装包里的一个Ruby脚本，所以需要先安装ruby.

[root@bfd-yiz-1p23 /opt]# yum -y install ruby rubygems
    [root@bfd-yiz-1p23 /opt]# gem install redis # 安装Redis客户端for Ruby

执行上述操作，可能会报下面的错误

ERROR:  Error installing redis:
        redis requires Ruby version >= 2.2.2.

解决方案：升级ruby的版本。

[root@bfd-yiz-1p23 /opt]# curl -L get.rvm.io | bash -s stable

这里要区分是否是root用户还是普通用户，因为不同用户下载到机器上的位置是不一样的。

root用户下载到的位置是/usr/local/rvm
    普通用户下载到的位置是/home/imoa/rvm

首先你会在下载的目录下找到rvm-1.29.4.tgz包，解压后，进行source 一下。

#到 /usr/local/rvm/archives目录下，要解压rvm-1.29.4包，再source
    root用户：source  /usr/local/rvm/archives/rvm-1.29.4/scripts/rvm

修改rvm的ruby安装源，修改 RVM 的 Ruby 安装源到 Ruby China 的 [Ruby 镜像服务器][Ruby]，这样能提高安装速度。

[root@bfd-yiz-1p23 /opt]# echo "ruby_url=https://cache.ruby-china.org/pub/ruby" >/usr/lcoal/rvm/user/db
    
    #如果报错说没有上述目录或者文件，则需要手动去创建
    mkdir -p  /usr/lcoal/rvm/user
    cd /usr/lcoal/rvm/user
    touch db

正常情况下，我们查看ruby的版本，然后进行安装

[root@bfd-yiz-1p23 /opt]# rvm list known  #查看版本
    [root@bfd-yiz-1p23 /opt]# rvm install 2.4.6  #安装2.4.6版本
    
    Searching for binary rubies, this might take some time.
    No binary rubies available for: centos/7/x86_64/ruby-2.4.6.
    Continuing with compilation. Please read 'rvm help mount' to get more information on binary rubies.
    Checking requirements for centos.
    Requirements installation successful.
    Installing Ruby from source to: /usr/local/rvm/archives/rvm-1.29.8/rubies/ruby-2.4.6, this may take a while depending on your cpu(s)...
    -bash: cd: /usr/local/rvm/archives/rvm-1.29.8/src: No such file or directory
    ruby-2.4.6 - #downloading ruby-2.4.6, this may take a while depending on your connection...
      % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                     Dload  Upload   Total   Spent    Left  Speed
    100 12.0M  100 12.0M    0     0  1239k      0  0:00:09  0:00:09 --:--:-- 1492k
    ruby-2.4.6 - #extracting ruby-2.4.6 to /usr/local/rvm/archives/rvm-1.29.8/src/ruby-2.4.6.....
    mv: cannot move ‘/usr/local/rvm/archives/rvm-1.29.8/tmp/rvm_src_6552/ruby-2.4.6’ to ‘/usr/local/rvm/archives/rvm-1.29.8/src/ruby-2.4.6’: No such file or directory
    -bash: cd: /usr/local/rvm/archives/rvm-1.29.8/src/ruby-2.4.6: No such file or directory
    
    
    [root@wyl01 archives]# mkdir -p /usr/local/rvm/archives/rvm-1.29.8/src/ruby-2.4.6

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70][]

在安装的过程中可能会报一些目录不存在的错误，根据提示，去创建这些目录即可。

Continuing with compilation. Please read 'rvm help mount' to get more information on binary rubies.
    df: ‘/usr/local/rvm/archives/rvm-1.29.5/rubies’: No such file or directory
    Not enough space (0MB) to install ruby (440MB).
    
    
    cd: /usr/local/rvm/archives/rvm-1.29.5/archives: No such file or directory

这里注意，在安装rvm的时候，会导致你的机器ctrl+c功能失效，可[参考该文章][Link 1]。其实卸载rvm应该就可以了。因为这个地方source一下rvm了，可能导致的。

ruby升级完后，不要忘了执行

$ gem install redis # 进行 Redis 接口安装

### 3.2创建Redis集群. ###

安装完ruby2.4.4后我们就可以进行以下操作, 创建Redis集群.

[root@bfd-yiz-1p23 /opt/redis/]# redis-trib.rb create --replicas 1 ${ip1}:7000 ${ip2}:7000 ${ip3}:7000 ${ip1}:7001 ${ip2}:7001 ${ip3}:7001
      
    >>> Creating cluster
    Can I set the above configuration? (type 'yes' to accept): yes       #输入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 127.0.0.1:7000)
    M: d5a834d075fd93eefab877c6ebb86efff680650f ${ip1}:7000
       slots:0-5460 (5461 slots) master
       1 additional replica(s)
    M: be2718476eba4e56f696e56b75e67df720b7fc24 ${ip2}:7000
       slots:10923-16383 (5461 slots) master
       1 additional replica(s)
    M: 13d0c397604a0b2644244c37b666fce83f29faa8 ${ip3}:7000
       slots:5461-10922 (5462 slots) master
       1 additional replica(s)
    S: 3d02f59b34047486faecc023685379de7b38076c ${ip1}:7001
       slots: (0 slots) slave
       replicates 13d0c397604a0b2644244c37b666fce83f29faa8
    S: 99c07119a449a703583019f7699e15afa0e41952 ${ip2}:7001
       slots: (0 slots) slave
       replicates d5a834d075fd93eefab877c6ebb86efff680650f
    S: dedf672f0a75faf37407ac4edd5da23bc4651e25 ${ip3}:7001
       slots: (0 slots) slave
       replicates be2718476eba4e56f696e56b75e67df720b7fc24
    [OK] All nodes agree about slots configuration.
    >>> Check for open slots...
    >>> Check slots coverage...
    [OK] All 16384 slots covered.
      
    根据提示,输入yes,我们会看到集群创建成功。

**报错1：\[ERR\] Sorry, can't connect to node 192.168.190.132:7000**

**解决方案：最容易忽略掉的一个原因是关闭防火墙和selinux.**

**报错2：重新搭建集群，可能出现的错误**

Can I set the above configuration? (type 'yes' to accept): yes
    /usr/local/rvm/archives/rvm-1.29.8/rubies/ruby-2.4.6/lib/ruby/gems/2.4.0/gems/redis-4.1.2/lib/redis/client.rb:126:in `call': ERR Slot 0 is already busy (Redis::CommandError)

**解决方案：重置集群信息**

在Redis集群中，清空各个节点上的数据，并执行cluster reset指令。

## 4.测试验证 ##

输入命令使用客户端`redis-cli`二进制访问某个实例，执行`set`和`get`的进行测试，如下所示：

$ redis-cli -h ${ip1}  -p 7000 -c

输入命令进行测试，如下图：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 1][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 2][]

## 5.redis密码认证 ##

在所有的redis实例中加上最下面的一句指令，并重启服务。

[root@wyl01 conf]# cat redis7000.conf 
    bind ${ip1}  
    port 7000
    cluster-enabled  yes
    cluster-config-file nodes.conf
    cluster-node-timeout 5000
    appendonly yes
    timeout 3600
    daemonize yes
      
    pidfile /opt/redis/pid/redis7000.pid
    logfile "/opt/redis/log/redis7000.log"
    dir "/opt/redis/data/7000"
      
    rename-command FLUSHALL ""
    rename-command FLUSHDB ""
    rename-command KEYS ""
    
    requirepass  "ifly"

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 3][]

redis-cli 连接集群的所有实例

redis-cli -u redis://192.168.52.128:7000,192.168.52.128:7001,192.168.52.129:7000,192.168.52.129:7001,192.168.52.130:7000,192.168.52.128:7001 -c -a ifly

[Ruby]: https://cache.ruby-china.org/
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70]: /images/20220411/0bcfe6575280401e8759a29625fa9829.png
[Link 1]: https://www.v2ex.com/t/489138
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 1]: /images/20220411/29c370b718414541a0a834c86ac5d107.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 2]: /images/20220411/5520660ef919432887443bf05992640d.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3d5bDk1Mjc_size_16_color_FFFFFF_t_70 3]: /images/20220411/a246afc886574fb682e6ea5da7764a52.png