redis集群搭建

环境准备

VM准备3台机器

centos7环境：
192.168.217.132   root root
192.168.217.133 root root
192.168.217.134 root root

单机安装

下载源码包(192.168.217.132)

cd /opt
wget https://download.redis.io/releases/redis-5.0.12.tar.gz

安装编译器(如已安装请忽略)

在安装redis之前先安装C++编译环境，查看目前服务器上gcc的版本：gcc -v
yum -y install gcc-c++

解压编译

tar -zxvf redis-5.0.12.tar.gz
指定redis安装目录为/usr/local/redis
cd redis-5.0.12 && make && make install PREFIX=/usr/local/redis
cp /opt/redis-5.0.12/*.conf /usr/local/redis

尝试运行

cd /usr/local/redis/bin
./redis-server ../redis.conf

建立软连接

添加到PATH
ln -s /usr/local/redis/bin/redis-cli /usr/local/bin/redis-cli
ln -s /usr/local/redis/bin/redis-server /usr/local/bin/redis-server
ln -s /usr/local/redis/bin/redis-sentinel /usr/local/bin/redis-sentinel
软连接建立路径必须都是绝对路径

启动服务

在任意路径都可以直接使用redis-cli和redis-server
redis-server /usr/local/redis/redis.conf

主从复制

节点非配

master    192.168.217.132
replica   192.168.217.133
replica   192.168.217.134

复制安装文件

scp -r /usr/local/redis root@192.168.217.133:/usr/local
scp -r /usr/local/redis root@192.168.217.134:/usr/local
# 需输入密码

修改redis.conf文件

# 后台启动
daemonize yes
protected-mode no
注释 bind 127.0.0.1
# 5.0版本以上使用的是replicaof，之前版本用savleof
# 配置从节点 从哪里复制数据，主节点不需要配置
replicaof 192.168.217.132 6379

如果主节点有密码校验，从节点需要配置masterauth
```
masterauth 123456
```

哨兵模式

修改三台机器上的sentinel.conf文件

daemonize yes
# 最后的2表示：至少有两个哨兵认为master节点挂了才进行选举
sentinel monitor mymaster 192.168.217.132 6379 2

启动哨兵

redis-sentinel /usr/local/redis/sentinel.conf

Cluster模式

准备配置文件(192.168.217.132)

mkdir /usr/local/redis/cluster
cp /opt/redis-5.0.12/redis.conf /usr/local/redis/cluster/redis_6379.conf
cp /opt/redis-5.0.12/redis.conf /usr/local/redis/cluster/redis_7379.conf

修改redis_6379.conf和redis_7379.conf

# 端口
port 6379 和 port 7379
# 后台运行
daemonize yes
# 进程号
pidfile /var/run/redis_6379.pid 和 pidfile /var/run/redis_7379.pid
# 工作目录
dir /usr/local/redis/cluster
# 开启集群
cluster-enabled yes
cluster-config-file nodes-6379.conf 和 cluster-config-file nodes-7379.conf

编写脚本/usr/local/redis/cluster/start.sh

#!/bin/bash
pid=`ps -ef | grep redis | grep -v grep | awk '{print $2}'`
for item in $pid
do
     kill -9 $item
done

redis-server /usr/local/redis/cluster/redis_6379.conf
redis-server /usr/local/redis/cluster/redis_7379.conf

远程复制

scp -r /usr/local/redis/cluster root@192.168.217.133:/usr/local/redis
scp -r /usr/local/redis/cluster root@192.168.217.134:/usr/local/redis

运行
```
cd /usr/local/redis/cluster
./start.sh
```

节点关联

任意一台机器
redis-cli --cluster create 192.168.217.132:6379 192.168.217.132:7379 192.168.217.133:6379 192.168.217.133:7379 192.168.217.134:6379 192.168.217.134:7379  --cluster-replicas 1

>>> Performing hash slots allocation on 6 nodes...
Master[0] -> Slots 0 - 5460
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
Adding replica 192.168.217.133:7379 to 192.168.217.132:6379
Adding replica 192.168.217.134:7379 to 192.168.217.133:6379
Adding replica 192.168.217.132:7379 to 192.168.217.134:6379
M: 6eb1b7c72b9bb0cfab13002156308cccbfedc723 192.168.217.132:6379
   slots:[0-5460] (5461 slots) master
S: 99c97e97efe3f4183aa9b2ccdc3f37602bfa4ce7 192.168.217.132:7379
   replicates 0d140c1a61c3588a453855ff60726d80adac1265
M: 640f4beb94a9d7f83a7f3efb0b6d9442db85e361 192.168.217.133:6379
   slots:[5461-10922] (5462 slots) master
S: c04ca1b7cf03c1bb574f63e9c766f1b6e233ad2e 192.168.217.133:7379
   replicates 6eb1b7c72b9bb0cfab13002156308cccbfedc723
M: 0d140c1a61c3588a453855ff60726d80adac1265 192.168.217.134:6379
   slots:[10923-16383] (5461 slots) master
S: dfbdf83174bd57ac615e62adec6006295932a9f0 192.168.217.134:7379
   replicates 640f4beb94a9d7f83a7f3efb0b6d9442db85e361
Can I set the above configuration? (type 'yes' to accept): yes
# 输入 yes

上面是随机分配主从节点、自动平均分配插槽，也可以自己定义

[root@node1 redis]# redis-cli -h 192.168.217.132
192.168.217.132:6379> cluster meet 192.168.217.133 6379
OK
192.168.217.132:6379> cluster meet 192.168.217.134 6379
OK
192.168.217.132:6379> cluster meet 192.168.217.132 7379
OK
192.168.217.132:6379> cluster meet 192.168.217.133 7379
OK
192.168.217.132:6379> cluster meet 192.168.217.134 7379
OK

查看集群节点

192.168.217.132:6379> cluster nodes
09ed6cdb5cb0f5ef706f18a488a8b23477f44e77 192.168.217.132:7379@17379 master - 0 1617472288407 1 connected
5cbc8237c517adafcb9d60c43dad34674cc54d1f 192.168.217.133:7379@17379 master - 0 1617472291496 0 connected
0e3808182116196c356bf8b79a605edc40156276 192.168.217.133:6379@16379 master - 0 1617472290464 4 connected
447b30e9ae3ba076e90e9080abb55f981d23dddb 192.168.217.132:6379@16379 myself,master - 0 1617472289000 2 connected
11c97e9c4db702681537cb02ede2bc4dab8a6ef4 192.168.217.134:7379@17379 master - 1617472293545 1617472289439 3 connected
1b968e634619001f65a807a47ec27cbb2a38480e 192.168.217.134:6379@16379 master - 0 1617472292521 5 connected
# 此时集群是不可用的，因为还没有分配槽

查看集群状态

192.168.217.132:6379> cluster info
cluster_state:fail
cluster_slots_assigned:0
cluster_slots_ok:0
cluster_slots_pfail:0
cluster_slots_fail:0
cluster_known_nodes:6
cluster_size:0
cluster_current_epoch:5
cluster_my_epoch:2
cluster_stats_messages_ping_sent:660
cluster_stats_messages_pong_sent:727
cluster_stats_messages_meet_sent:10
cluster_stats_messages_sent:1397
cluster_stats_messages_ping_received:727
cluster_stats_messages_pong_received:670
cluster_stats_messages_received:1397

指定副本节点

# 通过cluster replicate <master node_id> 指定从节点必须在从节点客户端进行，比如我现在需要把192.168.217.132:7379 实例指定为192.168.217.132:6379的从节点，我就必须在 redis-cli -h 192.168.217.132 -p 7379 去执行cluster replicate <master node_id>命令

192.168.217.132:7379> cluster replicate 447b30e9ae3ba076e90e9080abb55f981d23dddb
192.168.217.133:7379> cluster replicate 0e3808182116196c356bf8b79a605edc40156276
192.168.217.134:7379> cluster replicate 1b968e634619001f65a807a47ec27cbb2a38480e

# 结果
192.168.217.134:7379> cluster nodes
11c97e9c4db702681537cb02ede2bc4dab8a6ef4 192.168.217.134:7379@17379 myself,slave 1b968e634619001f65a807a47ec27cbb2a38480e 0 1616926898000 3 connected
0e3808182116196c356bf8b79a605edc40156276 192.168.217.133:6379@16379 master - 0 1616926905411 4 connected
1b968e634619001f65a807a47ec27cbb2a38480e 192.168.217.134:6379@16379 master - 0 1616926902357 5 connected
447b30e9ae3ba076e90e9080abb55f981d23dddb 192.168.217.132:6379@16379 master - 0 1616926903373 2 connected
09ed6cdb5cb0f5ef706f18a488a8b23477f44e77 192.168.217.132:7379@17379 slave 447b30e9ae3ba076e90e9080abb55f981d23dddb 0 1616926904393 2 connected
5cbc8237c517adafcb9d60c43dad34674cc54d1f 192.168.217.133:7379@17379 slave 0e3808182116196c356bf8b79a605edc40156276 0 1616926901337 4 connected

分配槽

[root@node1 redis]# redis-cli -h 192.168.217.132 -p 6379 cluster addslots {0..5460}
OK
[root@node1 redis]# redis-cli -h 192.168.217.133 -p 6379 cluster addslots {5461..10922}
OK
# 可以给副本节点分配槽
[root@node1 redis]# redis-cli -h 192.168.217.134 -p 7379 cluster addslots {10923..16383}
OK
# cluster nodes查看集群节点，发现134节点并没有成功分配槽
127.0.0.1:6379> cluster nodes
09ed6cdb5cb0f5ef706f18a488a8b23477f44e77 192.168.217.132:7379@17379 slave 447b30e9ae3ba076e90e9080abb55f981d23dddb 0 1617473587214 2 connected
5cbc8237c517adafcb9d60c43dad34674cc54d1f 192.168.217.133:7379@17379 slave 0e3808182116196c356bf8b79a605edc40156276 0 1617473586000 4 connected
0e3808182116196c356bf8b79a605edc40156276 192.168.217.133:6379@16379 master - 0 1617473588231 4 connected 5461-10922
447b30e9ae3ba076e90e9080abb55f981d23dddb 192.168.217.132:6379@16379 myself,master - 0 1617473587000 2 connected 0-5460
11c97e9c4db702681537cb02ede2bc4dab8a6ef4 192.168.217.134:7379@17379 slave 1b968e634619001f65a807a47ec27cbb2a38480e 0 1617473589251 5 connected
1b968e634619001f65a807a47ec27cbb2a38480e 192.168.217.134:6379@16379 master - 0 1617473590270 5 connected
# cluster info 查看集群状态，发现成功分配的槽数是10923个
127.0.0.1:6379> cluster info
cluster_state:fail
cluster_slots_assigned:10923
cluster_slots_ok:10923
cluster_slots_pfail:0
cluster_slots_fail:0
cluster_known_nodes:6
cluster_size:2
cluster_current_epoch:5
cluster_my_epoch:2
# 所以我个人猜测在从节点分配槽无效，没有论证
# 修改134分配槽
[root@node1 redis]# redis-cli -h 192.168.217.134 -p 6379 cluster addslots {10923..16383}

测试

# -c的作用是如果根据你的key计算出的hash slot不在0-5460间，会自动重定向到对应的redis-cli
redis-cli -c -h 192.168.217.132 -p 6379

192.168.217.132:6379> set name tom
-> Redirected to slot [5798] located at 192.168.217.133:6379
OK
192.168.217.133:6379>

最后编辑于：2021.04.04 16:04:55

人面猴
序言：七十年代末，一起剥皮案震惊了整个滨河市，随后出现的几起案子，更是在滨河造成了极大的恐慌，老刑警刘岩，带你破解...
沈念sama阅读 202,723评论 5赞 476
死咒
序言：滨河连续发生了三起死亡事件，死亡现场离奇诡异，居然都是意外死亡，警方通过查阅死者的电脑和手机，发现死者居然都...
沈念sama阅读 85,080评论 2赞 379
救了他两次的神仙让他今天三更去死
文/潘晓璐我一进店门，熙熙楼的掌柜王于贵愁眉苦脸地迎上来，“玉大人，你说我怎么就摊上这事。” “怎么了？”我有些...
开封第一讲书人阅读 149,604评论 0赞 335
道士缉凶录：失踪的卖姜人
文/不坏的土叔我叫张陵，是天一观的道长。经常有香客问我，道长，这世上最难降的妖魔是什么？我笑而不...
开封第一讲书人阅读 54,440评论 1赞 273
港岛之恋（遗憾婚礼）
正文为了忘掉前任，我火速办了婚礼，结果婚礼上，老公的妹妹穿的比我还像新娘。我一直安慰自己，他们只是感情好，可当我...
茶点故事阅读 63,431评论 5赞 364
恶毒庶女顶嫁案：这布局不是一般人想出来的
文/花漫我一把揭开白布。她就那样静静地躺着，像睡着了一般。火红的嫁衣衬着肌肤如雪。梳的纹丝不乱的头发上，一...
开封第一讲书人阅读 48,499评论 1赞 281
城市分裂传说
那天，我揣着相机与录音，去河边找鬼。笑死，一个胖子当着我的面吹牛，可吹牛的内容都是我干的。我是一名探鬼主播，决...
沈念sama阅读 37,893评论 3赞 395
双鸳鸯连环套：你想象不到人心有多黑
文/苍兰香墨我猛地睁开眼，长吁一口气：“原来是场噩梦啊……” “哼！你这毒妇竟也来了？” 一声冷哼从身侧响起，我...
开封第一讲书人阅读 36,541评论 0赞 256
万荣杀人案实录
序言：老挝万荣一对情侣失踪，失踪者是张志新（化名）和其女友刘颖，没想到半个月后，有当地人在树林里发现了一具尸体，经...
沈念sama阅读 40,751评论 1赞 296
护林员之死
正文独居荒郊野岭守林人离奇死亡，尸身上长有42处带血的脓包…… 初始之章·张勋以下内容为张勋视角年9月15日...
茶点故事阅读 35,547评论 2赞 319
白月光启示录
正文我和宋清朗相恋三年，在试婚纱的时候发现自己被绿了。大学时的朋友给我发了我未婚夫和他白月光在一起吃饭的照片。...
茶点故事阅读 37,619评论 1赞 329
活死人
序言：一个原本活蹦乱跳的男人离奇死亡，死状恐怖，灵堂内的尸体忽然破棺而出，到底是诈尸还是另有隐情，我是刑警宁泽，带...
沈念sama阅读 33,320评论 4赞 318
日本核电站爆炸内幕
正文年R本政府宣布，位于F岛的核电站，受9级特大地震影响，放射性物质发生泄漏。R本人自食恶果不足惜，却给世界环境...
茶点故事阅读 38,890评论 3赞 307
男人毒药：我在死后第九天来索命
文/蒙蒙一、第九天我趴在偏房一处隐蔽的房顶上张望。院中可真热闹，春花似锦、人声如沸。这庄子的主人今日做“春日...
开封第一讲书人阅读 29,896评论 0赞 19
一桩弑父案，背后竟有这般阴谋
文/苍兰香墨我抬头看了看天上的太阳。三九已至，却和暖如春，着一层夹袄步出监牢的瞬间，已是汗流浃背。一阵脚步声响...
开封第一讲书人阅读 31,137评论 1赞 259
情欲美人皮
我被黑心中介骗来泰国打工，没想到刚下飞机就差点儿被人妖公主榨干…… 1. 我叫王不留，地道东北人。一个月前我还...
沈念sama阅读 42,796评论 2赞 349
代替公主和亲
正文我出身青楼，却偏偏与公主长得像，于是被迫代替她去往敌国和亲。传闻我的和亲对象是个残疾皇子，可洞房花烛夜当晚...
茶点故事阅读 42,335评论 2赞 342