搭建Redis 集群

准备6台redis服务器，具体要求如下：
ip地址 端口 etho 日志文件名
192.168.4.51 6051 都可以接收连接请求 redis51.log
192.168.4.52 6052 都可以接收连接请求 redis52.log
192.168.4.53 6053 都可以接收连接请求 redis53.log
192.168.4.54 6054 都可以接收连接请求 redis54.log
192.168.4.55 6055 都可以接收连接请求 redis55.log
192.168.4.56 6056 都可以接收连接请求 redis56.log

环境准备
1 在6台redis服务器上运行服务，按照如下要求修改配置文件后，重启redis服务。

]#vim /etc/redis/6379.conf
bind 192.168.4.56 #只指定物理接口ip地址
port 6056 #不允许相同
daemonize yes
cluster-enabled yes
pidfile /var/run/redis_56.pid #不允许相同
cluster-config-file nodes-6056.conf #不允许相同
cluster-node-timeout 5000 #集群节点之间通信超时时间单位秒
:wq

停止服务
]# redis-cli -h 192.168.4.56 -p 6056 [ -a 密码 ] shutdown

启动服务
]# /etc/init.d/redis_6379 start

连接服务
[root@host56 ~]# redis-cli -h 192.168.4.56 -p 6056

装包： 在执行创建集群命令的主机安装ruby软件包（192.168.4.51）

安装解释ruby代码的软件包
[root@host51 ~]# yum -y install ruby rubygems
[root@host51 ~]# rpm -q rubygems ruby
rubygems-2.0.14.1-30.el7.noarch
ruby-2.0.0.648-30.el7.x86_64
[root@host51 ~]#

[root@host51 redis-cluster]# rpm -ivh --nodeps ruby-devel-2.0.0.648-30.el7.x86_64.rpm

安装ruby连接redis 接口程序 gem
[root@host51 redis-cluster]# which gem
/usr/bin/gem
[root@host51 redis-cluster]# gem install redis-3.2.1.gem
Successfully installed redis-3.2.1
Parsing documentation for redis-3.2.1
Installing ri documentation for redis-3.2.1
1 gem installed
[root@host51 redis-cluster]#

生成创建集群的脚本redis-trib.rb
[root@host51 src]# echo $PATH
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin

[root@host51 src]# cd
[root@host51 ~]# cp redis-4.0.8/src/redis-trib.rb /usr/local/bin/

创建集群：
[root@host51 ~]# redis-trib.rb create --replicas 1 \
192.168.4.51:6051 \
192.168.4.52:6052 \
192.168.4.53:6053 \
192.168.4.54:6054 \
192.168.4.55:6055 \
192.168.4.56:6056

[ERR] Node 192.168.4.54:6054 is not empty. Either the node already knows other nodes (check with

CLUSTER NODES) or contains some key in database 0.

[root@db54 ~]# redis-cli -h 192.168.4.54 -p 6054
192.168.4.54:6054> flushall
OK
192.168.4.54:6054> keys *
(empty list or set)
192.168.4.54:6054> flushall
OK
192.168.4.54:6054> save
OK
192.168.4.54:6054> quit
[root@db54 ~]#

槽位个数是16384
范围0-16383

[root@host51 ~]# redis-trib.rb create --replicas 1 192.168.4.51:6051 192.168.4.52:6052

192.168.4.53:6053 192.168.4.54:6054 192.168.4.55:6055 192.168.4.56:6056 #创建集群
--replicas 1 表示给master 配置1个slave

>> Creating cluster
>> Performing hash slots allocation on 6 nodes...
Using 3 masters:
192.168.4.51:6051
192.168.4.52:6052
192.168.4.53:6053
Adding replica 192.168.4.55:6055 to 192.168.4.51:6051
Adding replica 192.168.4.56:6056 to 192.168.4.52:6052
Adding replica 192.168.4.54:6054 to 192.168.4.53:6053
M: 0ec903c572270a90f3b140fba31aac15aaf5336b 192.168.4.51:6051
slots:0-5460 (5461 slots) master
M: 5278df7384edc9774b1a36b0b9d60a813a7424a9 192.168.4.52:6052
slots:5461-10922 (5462 slots) master
M: 6cdb4c64c48c0ee2ca35bf139660f31ca92821dc 192.168.4.53:6053
slots:10923-16383 (5461 slots) master
S: 388c33e7128fc961b381ad3b3c27c3c217912666 192.168.4.54:6054
replicates 6cdb4c64c48c0ee2ca35bf139660f31ca92821dc
S: 651f7d99965316c1b8a27a2e9b034a5b14c2be55 192.168.4.55:6055
replicates 0ec903c572270a90f3b140fba31aac15aaf5336b
S: a3af3096ee214c92a178eadf6e9299584899e62f 192.168.4.56:6056
replicates 5278df7384edc9774b1a36b0b9d60a813a7424a9
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 192.168.4.51:6051)
M: 0ec903c572270a90f3b140fba31aac15aaf5336b 192.168.4.51:6051
slots:0-5460 (5461 slots) master
1 additional replica(s)
M: 388c33e7128fc961b381ad3b3c27c3c217912666 192.168.4.54:6054
slots:9678 (1 slots) master
0 additional replica(s)
S: 651f7d99965316c1b8a27a2e9b034a5b14c2be55 192.168.4.55:6055
slots: (0 slots) slave
replicates 0ec903c572270a90f3b140fba31aac15aaf5336b
S: a3af3096ee214c92a178eadf6e9299584899e62f 192.168.4.56:6056
slots: (0 slots) slave
replicates 5278df7384edc9774b1a36b0b9d60a813a7424a9
M: 6cdb4c64c48c0ee2ca35bf139660f31ca92821dc 192.168.4.53:6053
slots:10923-16383 (5461 slots) master
0 additional replica(s)
M: 5278df7384edc9774b1a36b0b9d60a813a7424a9 192.168.4.52:6052
slots:5461-9677,9679-10922 (5461 slots) master
1 additional replica(s)
[OK] All nodes agree about slots configuration.
>> Check for open slots...
>> Check slots coverage...
[OK] All 16384 slots covered.
[root@host51 ~]#

[root@db52 redis-4.0.8]# cat /var/lib/redis/6052/nodes-6052.conf #集群配置文件
6cdb4c64c48c0ee2ca35bf139660f31ca92821dc 192.168.4.53:6053@16053 master - 0 1526096611374

3 connected 10923-16383
0ec903c572270a90f3b140fba31aac15aaf5336b 192.168.4.51:6051@16051 master - 0 1526096612000

1 connected 0-5460
388c33e7128fc961b381ad3b3c27c3c217912666 192.168.4.54:6054@16054 master - 0 1526096611043

4 connected 9678
a3af3096ee214c92a178eadf6e9299584899e62f 192.168.4.56:6056@16056 slave

5278df7384edc9774b1a36b0b9d60a813a7424a9 0 1526096611745 6 connected
651f7d99965316c1b8a27a2e9b034a5b14c2be55 192.168.4.55:6055@16055 slave

0ec903c572270a90f3b140fba31aac15aaf5336b 0 1526096613000 5 connected
5278df7384edc9774b1a36b0b9d60a813a7424a9 192.168.4.52:6052@16052 myself,master - 0

1526096612000 2 connected 5461-9677 9679-10922
vars currentEpoch 6 lastVoteEpoch 0
[root@db52 redis-4.0.8]#

set the above configuration? (type ‘yes‘ to accept): yes
/usr/local/share/gems/gems/redis-3.2.1/lib/redis/client.rb:113:in `call‘: ERR Slot 0 is already busy

(Redis::CommandError)
from /usr/local/share/gems/gems/redis-3.2.1/lib/redis.rb:2556:in block in method_missing‘<br/>from /usr/local/share/gems/gems/redis-3.2.1/lib/redis.rb:37:inblock in synchronize‘
from /usr/share/ruby/monitor.rb:211:in mon_synchronize‘<br/>from /usr/local/share/gems/gems/redis-3.2.1/lib/redis.rb:37:insynchronize‘
from /usr/local/share/gems/gems/redis-3.2.1/lib/redis.rb:2555:in method_missing‘<br/>from /usr/local/bin/redis-trib.rb:212:inflush_node_config‘
from /usr/local/bin/redis-trib.rb:906:in block in flush_nodes_config‘<br/>from /usr/local/bin/redis-trib.rb:905:ineach‘
from /usr/local/bin/redis-trib.rb:905:in flush_nodes_config‘<br/>from /usr/local/bin/redis-trib.rb:1426:increate_cluster_cmd‘
from /usr/local/bin/redis-trib.rb:1830:in `<main>‘
[root@mysql51 ~]#

[ERR] Node 192.168.4.53:6053 is not configured as a cluster node.

每台 redis服务 在本机登录 查看集群信息

[root@host51 ~]# redis-cli -c -h 192.168.4.51 -p 6051

-c 连接集群中的主机
[root@host51 ~]# redis-cli -c -h 192.168.4.51 -p 6051
192.168.4.51:6051> CLUSTER info
192.168.4.51:6051> CLUSTER nodes
192.168.4.51:6051> quit

192.168.4.51:6051> CLUSTER nodes
388c33e7128fc961b381ad3b3c27c3c217912666 192.168.4.54:6054@16054 master - 0 1526108837942

4 connected 9678
651f7d99965316c1b8a27a2e9b034a5b14c2be55 192.168.4.55:6055@16055 slave

0ec903c572270a90f3b140fba31aac15aaf5336b 0 1526108838444 5 connected
a3af3096ee214c92a178eadf6e9299584899e62f 192.168.4.56:6056@16056 slave

5278df7384edc9774b1a36b0b9d60a813a7424a9 0 1526108838946 6 connected
6cdb4c64c48c0ee2ca35bf139660f31ca92821dc 192.168.4.53:6053@16053 master - 0 1526108837942

3 connected 10923-16383
0ec903c572270a90f3b140fba31aac15aaf5336b 192.168.4.51:6051@16051 myself,master - 0

1526108836000 1 connected 0-5460
5278df7384edc9774b1a36b0b9d60a813a7424a9 192.168.4.52:6052@16052 master - 0

1526108839447 2 connected 5461-9677 9679-10922
192.168.4.51:6051>
192.168.4.51:6051>
192.168.4.51:6051>
192.168.4.51:6051> 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:4
cluster_current_epoch:6
cluster_my_epoch:1
cluster_stats_messages_ping_sent:26580
cluster_stats_messages_pong_sent:20656
cluster_stats_messages_update_sent:2
cluster_stats_messages_sent:47238
cluster_stats_messages_ping_received:20651
cluster_stats_messages_pong_received:26580
cluster_stats_messages_meet_received:5
cluster_stats_messages_received:47236
192.168.4.51:6051>

测试集群：连接master库 存储数据，对应的从库 会自动同步数据

[root@host51 ~]# redis-cli -c -h master库ip地址 -p 对应的端口号

[root@host51 ~]# redis-cli -c -h 192.168.4.51 -p 6051
192.168.4.51:6051> CLUSTER nodes
192.168.4.51:6051>

[root@host51 ~]# redis-cli -c -h 192.168.4.51 -p 6051
192.168.4.51:6051> set age 99
OK
192.168.4.51:6051> get age
"99"
192.168.4.51:6051>

[root@db54 ~]# redis-cli -c -h 192.168.4.54 -p 6054
192.168.4.54:6054> get age
-> Redirected to slot [741] located at 192.168.4.51:6051
"99"
192.168.4.51:6051>

二 、管理集群

2.1 集群选举测试 ： 停止某个主库的redi服务 对应的从会自动升级主库

2.2 向集群中添加新节点主机 192.168.4.77
1 在主机57运行redis服务（要做对应的集群配置）
装包 初始配置
修改配置文件
vim /etc/redis/6379.conf

停止服务
启动服务
查看端口号
11 yum -y install gcc
12 init 6
13 ls
14 tar -zxf redis-4.0.8.tar.gz
15 cd redis-4.0.8/
16 make
17 make install
18 ./utils/install_server.sh
19 /etc/init.d/redis_6379 status
20 /etc/init.d/redis_6379 stop
21 vim /etc/redis/6379.conf
22 /etc/init.d/redis_6379 start
23 /etc/init.d/redis_6379 status
[root@redis77 redis-4.0.8]# netstat -utnalp | grep redis-server
tcp 0 0 192.168.4.77:16077 0.0.0.0: LISTEN 4705/redis-server 1
tcp 0 0 192.168.4.77:6077 0.0.0.0: LISTEN 4705/redis-server 1
[root@redis77 redis-4.0.8]#

redis-trib.rb 选项

add-node 添加主机 (主机的角色是master)
check 检查集群
reshard 重新分配hash slot

2 把主机77添加到集群里
redis-trib.rb add-node 192.168.4.77:6077 192.168.4.71:6071

[root@redis71 ~]# redis-trib.rb add-node 192.168.4.77:6077 192.168.4.71:6071

>> Adding node 192.168.4.77:6077 to cluster 192.168.4.71:6071
>> Performing Cluster Check (using node 192.168.4.71:6071)
S: 3bfd47406253a25be9b13504ff0ecdd8524e9a9a 192.168.4.71:6071
slots: (0 slots) slave
replicates abb585e1803d756eb0b2d9a1b23790b6712ee7b2
M: cc0f13a4183eb06a771c5190319f7ac40626d04b 192.168.4.73:6073
slots:10923-16383 (5461 slots) master
1 additional replica(s)
S: 5172f3a9cbd4fe8a488a8d45bdf60ae46ac9f900 192.168.4.76:6076
slots: (0 slots) slave
replicates 9717d02408e45ad996dcdbde6401bf1319727e22
S: 30aac94cf32bd5d2446b584be5e7c6b0c07f7da0 192.168.4.74:6074
slots: (0 slots) slave
replicates cc0f13a4183eb06a771c5190319f7ac40626d04b
M: 9717d02408e45ad996dcdbde6401bf1319727e22 192.168.4.72:6072
slots:5461-10922 (5462 slots) master
1 additional replica(s)
M: abb585e1803d756eb0b2d9a1b23790b6712ee7b2 192.168.4.75:6075
slots:0-5460 (5461 slots) master
1 additional replica(s)
[OK] All nodes agree about slots configuration.
>> Check for open slots...
>> Check slots coverage...
[OK] All 16384 slots covered.
>> Send CLUSTER MEET to node 192.168.4.77:6077 to make it join the cluster.
[OK] New node added correctly.
[root@redis71 ~]#

redis-trib.rb check 192.168.4.77:6077

[root@redis71 ~]# redis-trib.rb check 192.168.4.77:6077

>> Performing Cluster Check (using node 192.168.4.77:6077)
M: 6cd754d126998090dcbe409315d7577bcd3b5605 192.168.4.77:6077
slots: (0 slots) master
0 additional replica(s)
S: 5172f3a9cbd4fe8a488a8d45bdf60ae46ac9f900 192.168.4.76:6076
slots: (0 slots) slave
replicates 9717d02408e45ad996dcdbde6401bf1319727e22
M: abb585e1803d756eb0b2d9a1b23790b6712ee7b2 192.168.4.75:6075
slots:0-5460 (5461 slots) master
1 additional replica(s)
S: 3bfd47406253a25be9b13504ff0ecdd8524e9a9a 192.168.4.71:6071
slots: (0 slots) slave
replicates abb585e1803d756eb0b2d9a1b23790b6712ee7b2
M: cc0f13a4183eb06a771c5190319f7ac40626d04b 192.168.4.73:6073
slots:10923-16383 (5461 slots) master
1 additional replica(s)
M: 9717d02408e45ad996dcdbde6401bf1319727e22 192.168.4.72:6072
slots:5461-10922 (5462 slots) master
1 additional replica(s)
S: 30aac94cf32bd5d2446b584be5e7c6b0c07f7da0 192.168.4.74:6074
slots: (0 slots) slave
replicates cc0f13a4183eb06a771c5190319f7ac40626d04b
[OK] All nodes agree about slots configuration.
>> Check for open slots...
>> Check slots coverage...
[OK] All 16384 slots covered.
[root@redis71 ~]#

手动重新分配hash slot
]# redis-trib.rb reshard 192.168.4.77:6077

4096 拿出多少个 hash slot 给主机 192.168.4.77

主机 192.168.4.77 id 值

all 从当前所有的master里 获取hash slot 个数

检查集群主机状态信息
[root@redis71 ~]# redis-trib.rb check 192.168.4.77:6077

>> Performing Cluster Check (using node 192.168.4.77:6077)
M: 6cd754d126998090dcbe409315d7577bcd3b5605 192.168.4.77:6077
slots:0-1364,5461-6826,10923-12287 (4096 slots) master
0 additional replica(s)

删除节点主机
1 删除master角色的节点主机
1.1 移除占用的hash槽（slot）
1.2 删除主机

]#redis-trib.rb reshard 192.168.4.51:6051
4096 移除hash槽的个数
主机id 目标主机id
主机id 源主机id
done 设置完毕
yes 提交

]# redis-trib.rb del-node 192.168.4.51:6051 8eecda17577349125df9a6fcc37107c6c5f9bdc5

2 删除slave角色的节点主机

]# redis-trib.rb del-node 192.168.4.51:6051 651f7d99965316c1b8a27a2e9b034a5b14c2be55

添加slave角色主机
1 主机运行redis服务 并设置集群配置
2 添加slave角色主机

]# redis-trib.rb add-node --slave 要添加到集群中主机ip：端口 集群中已有的任意一个主机的ip:端口

[ERR] Node 192.168.4.54:6054 is not empty. Either the node already knows other nodes (check with

CLUSTER NODES) or contains some key in database 0.

rm -rf /var/lib/redis/6379/
mkdir /var/lib/redis/6379/
chmod -R a+w /var/lib/redis/6379/
rm -rf /var/run/redis_6379.pid
/etc/init.d/redis_6379 start
211 /etc/init.d/redis_6379 status
212 redis-cli -h 192.168.4.54 -p 6054

[root@db54 redis]# redis-cli -c -h 192.168.4.54 -p 6054
192.168.4.54:6054> CLUSTER info
cluster_state:fail
cluster_slots_assigned:0
192.168.4.54:6054> CLUSTER RESET
OK

192.168.4.54:6054>quit;

]# redis-trib.rb add-node --slave 192.168.4.54:6054 192.168.4.51:6051

192.168.4.54:6054> CLUSTER nodes
5278df7384edc9774b1a36b0b9d60a813a7424a9 192.168.4.52:6052@16052 master - 0

1526121459000 2 connected 5461-9677 9679-10922
6cdb4c64c48c0ee2ca35bf139660f31ca92821dc 192.168.4.53:6053@16053 master - 0 1526121460637

3 connected 10923-16383
f7d5c6065d155b5d3d02a4efba10ca2d3722f634 192.168.4.54:6054@16054 myself,slave

0ec903c572270a90f3b140fba31aac15aaf5336b 0 1526121460000 0 connected
0ec903c572270a90f3b140fba31aac15aaf5336b 192.168.4.51:6051@16051 master - 0 1526121459635

8 connected 0-5460 9678
a3af3096ee214c92a178eadf6e9299584899e62f 192.168.4.56:6056@16056 slave

5278df7384edc9774b1a36b0b9d60a813a7424a9 0 1526121460000 2 connected
192.168.4.54:6054>

原文：http://blog.51cto.com/13773584/2134345