潇湘夜雨移动版

介绍：redis集群使用数据分片（sharding）而非一致性哈希（consistency hashing）来实现：

一个Redis集群包含16384个哈希槽（hash slot），数据库中的每个键都属于这16384个哈希槽

的其中一个，集群使用公式CRC16(key)%16384来计算键key属于那个槽，其中CRC16(key)语句用

于计算键key的CRC16校验和。

 

一、redis-cluster配置

 

1、安装ruby环境

用来执行redis-trib.rb脚本

[root@localhost ~]# yum install ruby rubygems -y

[root@localhost ~]# wget https://rubygems.org/downloads/redis-3.2.2.gem

 

[root@localhost ~]# gem install redis-3.2.2.gem 

Successfully installed redis-3.2.2

1 gem installed

Installing ri documentation for redis-3.2.2...

Installing RDoc documentation for redis-3.2.2...

 

 

2、 准备多个redis节点

 

配置文件：

[root@localhost redis]# cat redis3.conf 

bind 0.0.0.0

protected-mode no

port 6381

tcp-backlog 511

timeout 0

tcp-keepalive 300

daemonize yes

supervised no

pidfile "/var/run/redis_6381.pid" #自定义

loglevel notice

logfile "/var/log/redis/redis3.log" #自定义

databases 16

save 900 1

save 300 10

save 60 10000

stop-writes-on-bgsave-error yes

rdbcompression yes

rdbchecksum yes

dbfilename "dump3.rdb" #自定义

dir "/data/redis"

slave-serve-stale-data yes

slave-read-only yes

repl-diskless-sync no

repl-diskless-sync-delay 5

repl-disable-tcp-nodelay no

slave-priority 90

appendonly no

appendfilename "appendonly.aof"

appendfsync everysec

no-appendfsync-on-rewrite no

auto-aof-rewrite-percentage 100

auto-aof-rewrite-min-size 64mb

aof-load-truncated yes

lua-time-limit 5000

slowlog-log-slower-than 10000

slowlog-max-len 128

latency-monitor-threshold 0

notify-keyspace-events ""

hash-max-ziplist-entries 512

hash-max-ziplist-value 64

list-max-ziplist-size -2

list-compress-depth 0

set-max-intset-entries 512

zset-max-ziplist-entries 128

zset-max-ziplist-value 64

hll-sparse-max-bytes 3000

activerehashing yes

client-output-buffer-limit normal 0 0 0

client-output-buffer-limit slave 256mb 64mb 60

client-output-buffer-limit pubsub 32mb 8mb 60

hz 10

aof-rewrite-incremental-fsync yes

 

# Generated by CONFIG REWRITE

cluster-enabled yes #开启集群

cluster-config-file nodes6381.conf  #自定义

注意：自定义的部分，各个节点不能相同。

 

集群配置解析

 

默认情况下不能从slaves读取数据，但建立连接后，执行一次命令READONLY，该slaves即可读取数据。

否则只能以redis-cli -c -h -p命令登录 -c是以集群方式

cluster-enabled yes    集群开关，默认是不开启集群模式。

cluster-config-file nodes-6379.conf  集群配置文件的名称，每个节点都有一个集群相关的配置文件，持久化保存集群的信息。这个文件并不需要手动配置，这个配置文件有 Redis生成并更新，每个Redis集群节点需要一个单独的配置文件，请确保与实例运行的系统中配置文件名称不冲突。

cluster-node-timeout 15000 节点互连超时的阀值。集群节点超时毫秒数。即节点与集群其他节点断开多长时间将被认定为超时。建议稍微大一点

cluster-slave-validity-factor 10 在进行故障转移的时候，全部slave都会请求申请为master，但是有些slave可能与master断开连接一段时间了，导致数据过于陈旧，这样的slave不应该被提升为master。该参数就是用来判断slave节点与master断线的时间是否过长。判断方法是：比较slave断开连接的时间和(node-timeout * slave-validity-factor)+ repl-ping-slave-period如果节点超时时间为三十秒, 并且slave-validity-factor为10,假设默认的repl-ping-slave-period是10秒，即如果超过310秒slave将不会尝试进行故障转移

cluster-migration-barrier 1 master的slave数量大于该值，slave才能迁移到其他孤立master上，如这个参数若被设为2，那么只有当一个主节点拥有2个可工作的从节点时，它的一个从节点才会尝试迁移。

cluster-require-full-coverage yes 默认情况下，集群全部的slot有节点负责，集群状态才为ok，才能提供服务。设置为no，可以在slot没有全部分配的时候提供服务。不建议打开该配置，这样会造成分区的时候，小分区的master一直在接受写请求，而造成很长时间数据不一致。

/usr/local/redis/bin/redis-trib.rb --replicas 1 192.168.198.128:6379 192.168.198.128:6380 192.168.198.130:6379 192.168.198.130:6380

 

 

启动各个节点

 

3、创建集群

 

[root@localhost redis]# /usr/local/redis/bin/redis-trib.rb create --replicas 1 192.168.198.128:6379 192.168.198.128:6380 192.168.198.130:6379 192.168.198.130:6380

>>> Creating cluster

[ERR] Node 192.168.198.130:6379 is not empty. Either the node already knows other nodes (check with CLUSTER NODES) or contains some key in database 0.

解决方法：用 FLUSHALL，命令清除已经存在的键值。

[root@localhost redis]# /usr/local/redis/bin/redis-trib.rb create --replicas 1 192.168.198.128:6379 192.168.198.128:6380 192.168.198.130:6379 192.168.198.130:6380

>>> Creating cluster

*** ERROR: Invalid configuration for cluster creation.

*** Redis Cluster requires at least 3 master nodes.

*** This is not possible with 4 nodes and 1 replicas per node.

*** At least 6 nodes are required.

 

错误原因：redis cluster至少需要6个节点，至少3个主节点

 

/usr/local/redis/bin/redis-trib.rb create --replicas 1 192.168.198.128:6379 192.168.198.128:6380 192.168.198.128:6381 192.168.198.130:6379 192.168.198.130:6380 192.168.198.130:6381

 

[root@localhost ~]# /usr/local/redis/bin/redis-trib.rb create --replicas 1 192.168.198.128:6379 192.168.198.128:6380 192.168.198.128:6381 192.168.198.130:6379 192.168.198.130:6380 192.168.198.130:6381

>>> Creating cluster

>>> Performing hash slots allocation on 6 nodes...

Using 3 masters:

192.168.198.128:6379

192.168.198.130:6379

192.168.198.128:6380

Adding replica 192.168.198.130:6380 to 192.168.198.128:6379

Adding replica 192.168.198.128:6381 to 192.168.198.130:6379

Adding replica 192.168.198.130:6381 to 192.168.198.128:6380

M: 367f09037eb2aa0d5905bb826ee113458a965267 192.168.198.128:6379

  slots:0-5460,5907,12054 (5463 slots) master

M: 4eda271e009e929e47573efdce97f60b2ae127a3 192.168.198.128:6380

  slots:1559,2960,5907,10923-16383 (5464 slots) master

S: 64f4db57670eb12937aff6b24c7d074417f3a961 192.168.198.128:6381

  replicates 3858fa7dfd1255cddbb0d53320457138885c1790

M: 3858fa7dfd1255cddbb0d53320457138885c1790 192.168.198.130:6379

  slots:1559,2960,5461-10922,12054 (5465 slots) master

S: 3690d1f3e44fc1a331fd3c524ded78d0a078be97 192.168.198.130:6380

  replicates 367f09037eb2aa0d5905bb826ee113458a965267

S: 28132897c4ab835471a868084f123e7878258b05 192.168.198.130:6381

  replicates 4eda271e009e929e47573efdce97f60b2ae127a3

Can I set the above configuration? (type 'yes' to accept): yes

/usr/lib/ruby/gems/1.8/gems/redis-3.2.2/lib/redis/client.rb:114:in `call': ERR Slot 1559 is already busy (Redis::CommandError)

from /usr/lib/ruby/gems/1.8/gems/redis-3.2.2/lib/redis.rb:2646:in `method_missing'

from /usr/lib/ruby/gems/1.8/gems/redis-3.2.2/lib/redis.rb:57:in `synchronize'

from /usr/lib/ruby/1.8/monitor.rb:242:in `mon_synchronize'

from /usr/lib/ruby/gems/1.8/gems/redis-3.2.2/lib/redis.rb:57:in `synchronize'

from /usr/lib/ruby/gems/1.8/gems/redis-3.2.2/lib/redis.rb:2645:in `method_missing'

from /usr/local/redis/bin/redis-trib.rb:212:in `flush_node_config'

from /usr/local/redis/bin/redis-trib.rb:776:in `flush_nodes_config'

from /usr/local/redis/bin/redis-trib.rb:775:in `each'

from /usr/local/redis/bin/redis-trib.rb:775:in `flush_nodes_config'

from /usr/local/redis/bin/redis-trib.rb:1296:in `create_cluster_cmd' (责任编辑：liangzh)