记:redis集群安装

redis版本:redis-5.0.3
安装目录:/root/data/apps
集群目录:/root/data/apps/cluster
redis集群:至少需要3个主节点，每个主节点有一个从节点总共6个节点

• 解压到指定目录
  tar -zxvf redis-5.0.3.tar.gz -C /root/data/apps

• 安装redis于redis-5.0.3目录
  注：这里最好指定make目录 PREFIX=/root/data/apps/redis-5.0.3
  /root/data/apps/redis-5.0.3目录下执行：make PREFIX=/root/data/apps/redis-5.0.3 install

• 新建目录/root/data/apps/cluster

• 在cluster下新建6个子目录，分别为7001、7002、7003、7004、7005、7006

• 复制bin目录下的所有文件到cluster6个子目录下
  /root/data/apps目录下执行：cp -r redis-5.0.3/bin/* cluster/7001~7006/

• 复制redis配置文件到目录7001~7006
  /root/data/apps目录下执行：cp redis-5.0.3/redis.conf /root/data/apps/cluster/7001~7006/

• 修改配置文件redis.conf，相应目录对应相应端口
  port 7001
  daemonize yes
  pidfile ./redis_7001.pid
  cluster-enabled yes
  cluster-config-file nodes_7001.conf

• 分别启动6个节点的redis
  在/root/data/apps目录下执行 ./cluster/7001/redis-server ./cluster/7001/redis.conf
  在/root/data/apps目录下执行 ./cluster/7002/redis-server ./cluster/7002/redis.conf
  …
  …

• 查看6个节点是启动成
  ps -ef | grep redis

• redis集群依赖于ruby环境

• 安装ruby
  yum install ruby
  yum install rubygems

安装ruby后可能会出现版本过低现象，需要通过升级Ruby的版本来解决

• 执行集群启动命令任意节点下执行
  例：/root/data/apps/cluster/7001目录下执行 ./redis-cli --cluster create 127.0.0.1:7001 127.0.0.1:7002 127.0.0.1:7003 127.0.0.1:7004 127.0.0.1:7005 127.0.0.1:7006 --cluster-replicas 1

replicas指定为1表示每个主节点有一个从节点
注意：将127.0.0.1换为本机ip地址，避免报错Connection refused: no further information

-------------------报错-------------------
[ERR] Node 127.0.0.1:7001 is not configured as a cluster node.
-------------------原因-------------------
redis.conf 中的 #cluster-enabled yes 未打开

-------------------报错-------------------
16907:C 03 Apr 2019 11:04:10.170 # oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
16907:C 03 Apr 2019 11:04:10.170 # Redis version=5.0.4, bits=64, commit=00000000, modified=0, pid=16907, just started
16907:C 03 Apr 2019 11:04:10.170 # Configuration loaded
16908:M 03 Apr 2019 11:04:10.172 # Sorry, the cluster configuration file nodes.conf is already used by a different Redis Cluster node. Please make sure that different nodes use different cluster configuration files.
-------------------原因-------------------
redis.conf 中的 #cluster-config-file nodes_7001.conf 未打开

• 集群成功!

Performing Cluster Check (using node 47.9.2.11:7001)
M: 0a64cec2fa726fdcbfb02ffd99889c32f79e1d32 47.9.2.11:7001
slots:[0-5460] (5461 slots) master
1 additional replica(s)
S: 428b57bcf684d272a31b5054ff5bc23c67f070bc 47.9.2.11:7005
slots: (0 slots) slave
replicates 11260547eabb686b34cf2a77cdb79443ac0a4dee
S: 5a6ea80803efa6fd40b5677f9e30acef067119ce 47.9.2.11:7006
slots: (0 slots) slave
replicates 0a64cec2fa726fdcbfb02ffd99889c32f79e1d32
M: 11260547eabb686b34cf2a77cdb79443ac0a4dee 47.9.2.11:7003
slots:[10923-16383] (5461 slots) master
1 additional replica(s)
S: dc87f6c3e5ebd9e913b09fb03f387009f37be490 47.9.2.11:7004
slots: (0 slots) slave
replicates fb7f9e363268b6900ffd65e0c37b16a6f2ad3046
M: fb7f9e363268b6900ffd65e0c37b16a6f2ad3046 47.9.2.11:7002
slots:[5461-10922] (5462 slots) master
1 additional replica(s)
[OK] All nodes agree about slots configuration.

• 查看集群，验证是否成功
  /root/data/apps/cluster/7001目录下查看 : ./redis-cli -c -p 7001
  输入CLUSTER INFO
  查询结果如下：

127.0.0.1:7001> 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:3
cluster_current_epoch:6
cluster_my_epoch:1
cluster_stats_messages_ping_sent:1312
cluster_stats_messages_pong_sent:87
cluster_stats_messages_fail_sent:6
cluster_stats_messages_sent:1405
cluster_stats_messages_ping_received:87
cluster_stats_messages_pong_received:85
cluster_stats_messages_fail_received:2
cluster_stats_messages_received:174

• 添加外网访问
  改：
  bind 127.0.0.1
  protected-mode yes
  为：
  #bind 127.0.0.1
  protected-mode no

-------------------报错-------------------
spring boot集成报错Connection refused: no further information: /127.0.0.1:7003
-------------------原因-------------------
原因是集成时使用的是127.0.0.1创建的集成
./redis-cli --cluster create 127.0.0.1:7001 127.0.0.1:7002 127.0.0.1:7003 127.0.0.1:7004 127.0.0.1:7005 127.0.0.1:7006 --cluster-replicas 1
将127.0.0.1换为本机ip地址，再次创建集群。问题解决
源码分析是创建连接工厂的时候，只会生成一个正确的地址，然后使用创建集群的地址127.0.0.1

因集群已经创建，故需要删除生成的配置文件nodes.conf 和dump.rdb、appendonly.aof持久化文件文件

• 放开安全组
  创建集群一直等待 Waiting for the cluster to join
  因：
  redis集群不仅需要开通redis客户端连接的端口，而且需要开通集群总线端口
  集群总线端口为redis客户端连接的端口 + 10000
  如redis端口为6379
  则集群总线端口为16379
  故：
  放开安全组