Centos7.x离线安装redis-cluster集群

最新推荐文章于 2024-09-05 16:14:35 发布

zf_csdn

最新推荐文章于 2024-09-05 16:14:35 发布

阅读量1.4k

点赞数 1

分类专栏： Linux 文章标签： redis redis-cluster离线安装 ruby tcl8.6

本文链接：https://blog.csdn.net/zf_csdn/article/details/88304605

版权

Linux 专栏收录该内容

4 篇文章 0 订阅

订阅专栏

一、安装背景

由于工作需要，要求在测试环境安装redis-cluster集群，但是测试环境又不能访问外网，只能通过下载一些离线安装包进行安装，环境准备中提供所有离线安装所需的文件的下载地址，包括ruby-2.4.1.tar.gz，rubygems-2.7.6.tgz，zlib-1.2.11.tar.gz，redis-4.0.2.gem，tcl8.6.8-src.tar.gz

二、环境准备

系统版本 CentOS 7.2
软件版本 redis-4.0.2.tar.gz
所需的依赖包

通过 cat /etc/redhat-release 可查看系统版本

# cat /etc/redhat-release 
CentOS Linux release 7.2.1511 (Core)

三、开始安装

1. 离线安装ruby

将ruby-2.4.1.tar.gz上传至服务器，解压安装包

# tar -zxvf ruby-2.4.1.tar.gz

进入ruby-2.4.1文件，并配置安装目录

# cd ruby-2.4.1
# ./configure --prefix=/usr/local/ruby

通过echo $?命令查看是否执行完成，0表示完成，其他值为未完成

# echo $?   
0

开始编译安装

# make && make install

安装完成后，将ruby加入到系统路径里，通过source命令让/etc/profile立即生效

# echo "PATH=$PATH:/usr/local/ruby/bin" >> /etc/profile
# source /etc/profile

为ruby创建软连接

# which ruby
/usr/local/ruby/bin/ruby
# rm -rf /usr/bin/ruby
# ln -s /usr/local/ruby/bin/ruby /usr/bin/ruby
# which ruby
/usr/bin/ruby

此时ruby便安装完成，可通过ruby -v查看版本

# ruby -v
ruby 2.4.1p111 (2017-03-22 revision 58053) [x86_64-linux]

2. 离线安装rubygems

将rubygems-2.7.6.tgz上传至服务器，解压

# tar -xvf rubygems-2.7.6.tgz
# chown -R root.root rubygems-2.7.6
# mv rubygems-2.7.6 /usr/local/

安装rubygems

# cd /usr/local/rubygems-2.7.6/
# ruby setup.rb

此时可能会出现以下报错

# ruby setup.rb
Bundler 1.16.1 installed
RubyGems 2.7.6 installed
/usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require': cannot load such file -- zlib (LoadError)
	from /usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require'
	from /usr/local/rubygems-2.7.6/lib/rubygems/package.rb:47:in `<top (required)>'
	from /usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require'
	from /usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require'
	from /usr/local/rubygems-2.7.6/lib/rubygems/commands/pristine_command.rb:3:in `<top (required)>'
	from /usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require'
	from /usr/local/rubygems-2.7.6/lib/rubygems/core_ext/kernel_require.rb:59:in `require'
	from /usr/local/rubygems-2.7.6/lib/rubygems/commands/setup_command.rb:583:in `regenerate_binstubs'
	from /usr/local/rubygems-2.7.6/lib/rubygems/commands/setup_command.rb:155:in `execute'
	from /usr/local/rubygems-2.7.6/lib/rubygems/command.rb:313:in `invoke_with_build_args'
	from /usr/local/rubygems-2.7.6/lib/rubygems/command_manager.rb:171:in `process_args'
	from /usr/local/rubygems-2.7.6/lib/rubygems/command_manager.rb:141:in `run'
	from /usr/local/rubygems-2.7.6/lib/rubygems/gem_runner.rb:59:in `run'
	from setup.rb:46:in `<main>'

从报错信息来看，大概意思是无法加载zlib文件，怀疑是ruby在编译安装的时候没有编译这个zlib，于是决定把ruby重新编译安装。

# cd /usr/local/src/redis-cluster/ruby-2.4.1/ext/zlib/
# ruby extconf.rb
# make

/usr/local/src/redis-cluster/ruby-2.4.1/为reby的路径，当执行make命令时，可能又会报以下错误

# make
make: *** No rule to make target `/include/ruby.h', needed by `zlib.o'.  Stop.

此时，可修改当前目录下的Makefile文件

# vi Makefile
#zlib.o: $(top_srcdir)/include/ruby.h #把这一行替换成下面一行(278行)
zlib.o: ../../include/ruby.h

在这里插入图片描述

重新编译，即可成功

# make
# make install

此时zlib便加载成功了，然后在回到rubygems继续安装

# cd /usr/local/rubygems-2.7.6/
# ruby setup.rb

这时，应该是可以成功安装完成的。然后设置系统路径

# echo "PATH=$PATH:/usr/local/rubygems-2.7.6" >> /etc/profile
# source /etc/profile

这样，rubygems便是安装完成了，可查看当前版本

# gem -v
2.7.6

3. 离线配置rubygems的redisapi

将redis-4.0.2.gem上传至服务器，可先查看系统是否已经安装过

# gem list redis

*** LOCAL GEMS ***

我这里是没有安装过，执行安装命令

# gem install -l redis-4.0.2.gem
Successfully installed redis-4.0.2
Parsing documentation for redis-4.0.2
Installing ri documentation for redis-4.0.2
Done installing documentation for redis after 1 seconds
1 gem installed

这个便安装成功了，查看详情

# gem list redis

*** LOCAL GEMS ***

redis (4.0.2)

4. 离线安装tcl

上传tcl8.6.8-src.tar.gz到服务器，解压，指定安装路径，安装

# tcl8.6.8-src.tar.gz
# cd tcl8.6.8/unix/
# ./configure --prefix=/usr/local/tcl
# echo $?
# make 
# make install
# echo $?
# make install-private-headers

设置软连接，赋权

# ln -v -sf tclsh8.6 /usr/local/tcl/bin/tclsh
# chmod -v 755 /usr/local/tcl/lib/libtcl8.6.so
# echo "PATH=$PATH:/usr/local/tcl/bin" >> /etc/profile
# source /etc/profile

5. 离线安装redis

上传redis-4.0.2.tar.gz至服务器，解压，安装

# tar -zxf redis-4.0.2.tar.gz
# cd redis-4.0.2
# make
# echo $?

至此，redis变安装成功了，接下来我们便开始部署集群,由于redis-cluster至少需要3个主节点，若每个主节点配置一个从节点，则至少需要六个，因此我们在这里创建6个文件夹，当然，如果有多台服务器，可分布在各台服务器上

# mkdir /usr/local/src/redis-cluster/redis-4.0.2/redis-cluster
# cd /usr/local/src/redis-cluster/redis-4.0.2/redis-cluster
# mkdir {6500,6501,6502,6503,6504,6505}

然后，为每个节点创建配置文件，以6500为例，部署172.130.132.128:6500

# cat <<EOF>./6500/redis.conf
> port 6500
> bind 172.130.132.128
> daemonize yes
> pidfile /var/run/redis_6500.pid
> cluster-enabled yes
> cluster-config-file nodes_6500.conf
> cluster-node-timeout 10100
> appendonly yes
> EOF

同理，部署172.130.132.128:6501,只需修改下端口即可

# cat <<EOF>./6500/redis.conf
> port 6501
> bind 172.130.132.128
> daemonize yes
> pidfile /var/run/redis_6501.pid
> cluster-enabled yes
> cluster-config-file nodes_6501.conf
> cluster-node-timeout 10100
> appendonly yes
> EOF

以此类推，依次创建6502，6503，6504，6505的配置文件，配置完成后，启动所有节点

[root@i-epr36ck8 redis-cluster]# for((i=0;i<=5;i++)); do /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server /usr/local/src/redis-cluster/redis-4.0.2/redis-cluster/650$i/redis.conf; done
e20304:C 06 Mar 16:45:31.488 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20304:C 06 Mar 16:45:31.488 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20304, just started
20304:C 06 Mar 16:45:31.488 # Configuration loaded
20306:C 06 Mar 16:45:31.510 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20306:C 06 Mar 16:45:31.510 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20306, just started
20306:C 06 Mar 16:45:31.510 # Configuration loaded
20308:C 06 Mar 16:45:31.515 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20308:C 06 Mar 16:45:31.515 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20308, just started
20308:C 06 Mar 16:45:31.515 # Configuration loaded
20313:C 06 Mar 16:45:31.520 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20313:C 06 Mar 16:45:31.520 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20313, just started
20313:C 06 Mar 16:45:31.520 # Configuration loaded
20318:C 06 Mar 16:45:31.545 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20318:C 06 Mar 16:45:31.545 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20318, just started
20318:C 06 Mar 16:45:31.545 # Configuration loaded
20323:C 06 Mar 16:45:31.559 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
20323:C 06 Mar 16:45:31.559 # Redis version=4.0.2, bits=64, commit=00000000, modified=0, pid=20323, just started
20323:C 06 Mar 16:45:31.559 # Configuration loaded

查看状态

[root@i-epr36ck8 redis-cluster]# ps -ef |grep redis-server
root     20305     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6500 [cluster]
root     20307     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6501 [cluster]
root     20312     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6502 [cluster]
root     20317     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6503 [cluster]
root     20319     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6504 [cluster]
root     20324     1  0 16:45 ?        00:00:00 /usr/local/src/redis-cluster/redis-4.0.2/src/redis-server 172.130.132.128:6505 [cluster]
root     20742 17205  0 16:52 pts/0    00:00:00 grep --color=auto redis-server

redis.conf的配置说明

port 6500 配置集群的端口
bind 本机的IP 这里的默认配置是127.0.0.1改为内网ip
daemonsize yes 允许redis在后台运行
pidfile /var/run/redis_6500.pidbind 本机的IP 改成和端口一致
cluster-enabled yes 开启集群把注释去掉
cluster-config-file node_6500.conf 集群的配置，和端口一致
cluster-node-timeout 15000 请求超时，默认为15秒
appendonly yes aof日志开启，有需要就开启，每一次写操作都会记录一条日志。

压轴戏来了，创建集群

# cd /usr/local/src/redis-cluster/redis-4.0.2/src/
# ./redis-trib.rb create --replicas 1 172.130.132.128:6500 172.130.132.128:6501 172.130.132.128:6502 172.130.132.128:6503 172.130.132.
128:6504 172.130.132.128:6505

可能会出现以下报错（若无此错误，可忽略）

*** ERROR: Invalid configuration for cluster creation.
*** Redis Cluster requires at least 3 master nodes.
*** This is not possible with 5 nodes and 172 replicas per node.
*** At least 519 nodes are required.

于是，使用/redis-trib.rb check检查节点

[root@i-epr36ck8 src]# ./redis-trib.rb check 172.130.132.128:6500
>>> Performing Cluster Check (using node 172.130.132.128:6500)
M: 7e26d3b838e91ab5e6424ddd857d5593b6a1bfa1 172.130.132.128:6500
   slots: (0 slots) master
   0 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[ERR] Not all 16384 slots are covered by nodes.

这个往往是由于主node移除了，但是并没有移除node上面的slot，从而导致了slot总数没有达到16384，其实也就是slots分布不正确。所以在删除节点的时候一定要注意删除的是否是Master主节点。

此时，可使用./redis-trib.rb fix来修复每个端口

# ./redis-trib.rb fix 172.130.132.128:6500
# ./redis-trib.rb fix 172.130.132.128:6501
# ./redis-trib.rb fix 172.130.132.128:6502
# ./redis-trib.rb fix 172.130.132.128:6503
# ./redis-trib.rb fix 172.130.132.128:6504
# ./redis-trib.rb fix 172.130.132.128:6505

如果出现了以下错误，

/usr/local/ruby/lib/ruby/gems/2.4.0/gems/redis-4.0.2/lib/redis/client.rb:119:in `call': ERR Slot 0 is already busy (Redis::CommandError)
	from /usr/local/ruby/lib/ruby/gems/2.4.0/gems/redis-4.0.2/lib/redis.rb:2854:in `block in method_missing'
	from /usr/local/ruby/lib/ruby/gems/2.4.0/gems/redis-4.0.2/lib/redis.rb:45:in `block in synchronize'
	from /usr/local/ruby/lib/ruby/2.4.0/monitor.rb:214:in `mon_synchronize'
	from /usr/local/ruby/lib/ruby/gems/2.4.0/gems/redis-4.0.2/lib/redis.rb:45:in `synchronize'
	from /usr/local/ruby/lib/ruby/gems/2.4.0/gems/redis-4.0.2/lib/redis.rb:2853:in `method_missing'
	from ./redis-trib.rb:212:in `flush_node_config'
	from ./redis-trib.rb:776:in `block in flush_nodes_config'
	from ./redis-trib.rb:775:in `each'
	from ./redis-trib.rb:775:in `flush_nodes_config'
	from ./redis-trib.rb:1296:in `create_cluster_cmd'
	from ./redis-trib.rb:1700:in `<main>'

错误提示是，
slot插槽被占用了（这是搭建集群前时，以前redis的旧数据和配置信息没有清理干净。）

用redis-cli 登录到每个节点执行 flushall 和 cluster reset 就可以了。

[root@i-epr36ck8 src]# ./redis-cli -h 172.130.132.128 -c -p 6500
172.130.132.128:6500> flushall
OK
172.130.132.128:6500> cluster reset
OK

再次执行集群创建命令，即可安装完成

[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

四、验证集群

# cd /usr/local/src/redis-cluster/redis-4.0.2/src
# ./redis-cli -h 172.130.132.128 -c -p 6500
172.130.132.128:6500> set test 1111
ok
172.130.132.128:6500> get test
"1111"
172.130.132.128:6500> quit
# ./redis-cli -h 172.130.132.128 -c -p 6501
172.130.132.128:6501> get test
"1111"

大功告成，完美收工~