Redis Cluster集群

1    Redis Cluster集群

redis3.0以后推出的redis cluster 集群方案,redis cluster集群保证了高可用、高性能、高可扩展性。

1.1   redis-cluster架构图

 

架构细节:

(1)所有的redis节点彼此互联(PING-PONG机制),内部使用二进制协议优化传输速度和带宽.

(2)节点的fail是通过集群中超过半数的节点检测失效时才生效.

(3)客户端与redis节点直连,不需要中间proxy.客户端不需要连接集群所有节点,连接集群中任何一个可用节点即可

(4)redis-cluster把所有的物理节点映射到[0-16383]slot,cluster 负责维护node<->slot<->value

Redis 集群中内置了 16384 个哈希槽,当需要在 Redis 集群中放置一个 key-value 时,redis 先对 key 使用 crc16 算法算出一个结果,然后把结果对 16384 求余数,这样每个 key 都会对应一个编号在 0-16383 之间的哈希槽,redis 会根据节点数量大致均等的将哈希槽映射到不同的节点

 

示例如下:

 

 

 

1.2   redis-cluster投票:容错

 

 

 

最小节点数:3

 

(1)节点失效判断:集群中所有master参与投票,如果半数以上master节点与其中一个master节点通信超过(cluster-node-timeout),认为该master节点挂掉.

 

(2)集群失效判断:什么时候整个集群不可用(cluster_state:fail)? 

 

Ø  如果集群任意master挂掉,且当前master没有slave,则集群进入fail状态。也可以理解成集群的[0-16383]slot映射不完全时进入fail状态。

 

Ø  如果集群超过半数以上master挂掉,无论是否有slave,集群进入fail状态。

 

 

 

1.3  安装Ruby环境

 

redis集群需要使用集群管理脚本redis-trib.rb,它的执行相应依赖ruby环境。

 

 

 

l  第一步:安装ruby

 

yum install ruby

yum install rubygems

 

l  第三步:安装rubyredis的接口程序redis-3.2.2.gem

 

gem install redis -V 3.2.2

 

l  第四步:复制redis-3.2.9/src/redis-trib.rb文件到/usr/local/redis目录

 

cp redis-3.2.9/src/redis-trib.rb /usr/local/redis-cluster/ -r

 

 

 

1.4   安装Redis集群(RedisCluster

 

Redis集群最少需要三台主服务器,三台从服务器

 

端口号分别为:7001~7006

 

 

 

l  第一步:创建7001实例,并编辑redis.conf文件,修改port7001

 

注意:创建实例,即拷贝单机版安装时,生成的bin目录,为7001目录。

 

 

 

 

 

l  第二步:修改redis.conf配置文件,打开Cluster-enable yes

 

 

 

 

 

 

 

l  第三步:复制7001,创建7002~7006实例,注意端口修改

 

l  第四步:启动所有的实例

 

l  第五步:创建Redis集群

 

./redis-trib.rb create --replicas 1 192.168.10.133:7001 192.168.10.133:7002 192.168.10.133:7003 192.168.10.133:7004 192.168.10.133:7005  192.168.10.133:7006

>>> Creating cluster

Connecting to node 192.168.10.133:7001: OK

Connecting to node 192.168.10.133:7002: OK

Connecting to node 192.168.10.133:7003: OK

Connecting to node 192.168.10.133:7004: OK

Connecting to node 192.168.10.133:7005: OK

Connecting to node 192.168.10.133:7006: OK

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

Using 3 masters:

192.168.10.133:7001

192.168.10.133:7002

192.168.10.133:7003

Adding replica 192.168.10.133:7004 to 192.168.10.133:7001

Adding replica 192.168.10.133:7005 to 192.168.10.133:7002

Adding replica 192.168.10.133:7006 to 192.168.10.133:7003

M: d8f6a0e3192c905f0aad411946f3ef9305350420 192.168.10.133:7001

   slots:0-5460 (5461 slots) master

M: 7a12bc730ddc939c84a156f276c446c28acf798c 192.168.10.133:7002

   slots:5461-10922 (5462 slots) master

M: 93f73d2424a796657948c660928b71edd3db881f 192.168.10.133:7003

   slots:10923-16383 (5461 slots) master

S: f79802d3da6b58ef6f9f30c903db7b2f79664e61 192.168.10.133:7004

   replicates d8f6a0e3192c905f0aad411946f3ef9305350420

S: 0bc78702413eb88eb6d7982833a6e040c6af05be 192.168.10.133:7005

   replicates 7a12bc730ddc939c84a156f276c446c28acf798c

S: 4170a68ba6b7757e914056e2857bb84c5e10950e 192.168.10.133:7006

   replicates 93f73d2424a796657948c660928b71edd3db881f

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 192.168.10.133:7001)

M: d8f6a0e3192c905f0aad411946f3ef9305350420 192.168.10.133:7001

   slots:0-5460 (5461 slots) master

M: 7a12bc730ddc939c84a156f276c446c28acf798c 192.168.10.133:7002

   slots:5461-10922 (5462 slots) master

M: 93f73d2424a796657948c660928b71edd3db881f 192.168.10.133:7003

   slots:10923-16383 (5461 slots) master

M: f79802d3da6b58ef6f9f30c903db7b2f79664e61 192.168.10.133:7004

   slots: (0 slots) master

   replicates d8f6a0e3192c905f0aad411946f3ef9305350420

M: 0bc78702413eb88eb6d7982833a6e040c6af05be 192.168.10.133:7005

   slots: (0 slots) master

   replicates 7a12bc730ddc939c84a156f276c446c28acf798c

M: 4170a68ba6b7757e914056e2857bb84c5e10950e 192.168.10.133:7006

   slots: (0 slots) master

   replicates 93f73d2424a796657948c660928b71edd3db881f

[OK] All nodes agree about slots configuration.

>>> Check for open slots...

>>> Check slots coverage...

[OK] All 16384 slots covered.

[root@localhost-0723 redis]#

 

 

 

1.5   命令客户端连接集群

 

命令:./redis-cli –h 127.0.0.1 –p 7001 –c

 

注意:-c 表示是以redis集群方式进行连接

 

./redis-cli -p 7006 -c

127.0.0.1:7006> set key1 123

-> Redirected to slot [9189] located at 127.0.0.1:7002

OK

127.0.0.1:7002>

 

 

 

1.6   查看集群的命令

 

Ø  查看集群状态

 

127.0.0.1:7003> 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:3

cluster_stats_messages_sent:926

cluster_stats_messages_received:926

 

 

 

Ø  查看集群中的节点:

 

 

 

127.0.0.1:7003> cluster nodes

7a12bc730ddc939c84a156f276c446c28acf798c 127.0.0.1:7002 master - 0 1443601739754 2 connected 5461-10922

93f73d2424a796657948c660928b71edd3db881f 127.0.0.1:7003 myself,master - 0 0 3 connected 10923-16383

d8f6a0e3192c905f0aad411946f3ef9305350420 127.0.0.1:7001 master - 0 1443601741267 1 connected 0-5460

4170a68ba6b7757e914056e2857bb84c5e10950e 127.0.0.1:7006 slave 93f73d2424a796657948c660928b71edd3db881f 0 1443601739250 6 connected

f79802d3da6b58ef6f9f30c903db7b2f79664e61 127.0.0.1:7004 slave d8f6a0e3192c905f0aad411946f3ef9305350420 0 1443601742277 4 connected

0bc78702413eb88eb6d7982833a6e040c6af05be 127.0.0.1:7005 slave 7a12bc730ddc939c84a156f276c446c28acf798c 0 1443601740259 5 connected

127.0.0.1:7003>

 

 

 

1.7   维护节点

 

集群创建成功后可以继续向集群中添加节点

 

1.7.1添加主节点

 

l  先创建7007节点

 

l  添加7007结点作为新节点

 

执行命令:./redis-trib.rb add-node 127.0.0.1:7007 127.0.0.1:7001

 

 

 

l  查看集群结点发现7007已添加到集群中

 

 

 

 

1.7.2hash槽重新分配(数据迁移)

 

添加完主节点需要对主节点进行hash槽分配,这样该主节才可以存储数据

 

 

 

l  查看集群中槽占用情况

 

 

 

redis集群有16384个槽,集群中的每个结点分配自已槽,通过查看集群结点可以看到槽占用情况。

 

 

 

l  给刚添加的7007结点分配槽

 

* 第一步:连接上集群(连接集群中任意一个可用结点都行)

 

./redis-trib.rb reshard 192.168.10.133:7001

 

 

 

* 第二步:输入要分配的槽数量

 

 

 

 

 

输入:3000,表示要给目标节点分配3000个槽

 

 

 

* 第三步:输入接收槽的结点id

 

 

 

输入:15b809eadae88955e36bcdbb8144f61bbbaf38fb

 

PS:这里准备给7007分配槽,通过cluster nodes查看7007结点id为:

 

15b809eadae88955e36bcdbb8144f61bbbaf38fb

 

 

 

* 第四步:输入源结点id

 

 

 

输入:all

 

 

 

* 第五步:输入yes开始移动槽到目标结点id

 

 

 

输入:yes

 

 

 

1.7.3添加从节点

 

 

 

l  添加7008从结点,将7008作为7007的从结点

 

 

 

命令:./redis-trib.rb add-node --slave --master-id  主节点id   新节点的ip和端口   旧节点ip和端口(集群中任一节点都可以)

 

 

 

执行如下命令:

 

./redis-trib.rb add-node --slave --master-id  35da64607a02c9159334a19164e68dd95a3b943c 192.168.10.103:7008 192.168.10.103:7001

 

 

 

35da64607a02c9159334a19164e68dd95a3b943c7007结点的id,可通过cluster nodes查看。

 

 

 

注意:如果原来该结点在集群中的配置信息已经生成到cluster-config-file指定的配置文件中(如果cluster-config-file没有指定则默认为nodes.conf),这时可能会报错:

 

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

 

解决方法是删除生成的配置文件nodes.conf,删除后再执行./redis-trib.rb add-node指令

 

 

 

l  查看集群中的结点,刚添加的70087007的从节点:

 

  

 

1.7.4删除结点

 

命令:./redis-trib.rb del-node 127.0.0.1:7005 4b45eb75c8b428fbd77ab979b85080146a9bc017

 

 

 

删除已经占有hash槽的结点会失败,报错如下:

 

[ERR] Node 127.0.0.1:7005 is not empty! Reshard data away and try again.

 

 

 

需要将该结点占用的hash槽分配出去(参考hash槽重新分配章节)。

 

1.8   Jedis连接集群

 

需要开启防火墙,或者直接关闭防火墙。

 

service iptables stop

 

 

 

1.8.1代码实现

 

创建JedisCluster类连接redis集群。

 

 

 

@Test

publicvoid testJedisCluster() throws Exception {

         //创建一连接,JedisCluster对象,在系统中是单例存在

         Set<HostAndPort> nodes = new HashSet<>();

         nodes.add(new HostAndPort("192.168.10.133", 7001));

         nodes.add(new HostAndPort("192.168.10.133", 7002));

         nodes.add(new HostAndPort("192.168.10.133", 7003));

         nodes.add(new HostAndPort("192.168.10.133", 7004));

         nodes.add(new HostAndPort("192.168.10.133", 7005));

         nodes.add(new HostAndPort("192.168.10.133", 7006));

         JedisCluster cluster = new JedisCluster(nodes);

         //执行JedisCluster对象中的方法,方法和redis一一对应。

         cluster.set("cluster-test", "my jedis cluster test");

         String result = cluster.get("cluster-test");

         System.out.println(result);

         //程序结束时需要关闭JedisCluster对象

         cluster.close();

}

 

1.8.2使用spring

 

Ø  配置applicationContext.xml

 

<!-- 连接池配置 -->

<bean id="jedisPoolConfig" class="redis.clients.jedis.JedisPoolConfig">

         <!-- 最大连接数 -->

         <property name="maxTotal" value="30" />

         <!-- 最大空闲连接数 -->

         <property name="maxIdle" value="10" />

         <!-- 每次释放连接的最大数目 -->

         <property name="numTestsPerEvictionRun" value="1024" />

         <!-- 释放连接的扫描间隔(毫秒) -->

         <property name="timeBetweenEvictionRunsMillis" value="30000" />

         <!-- 连接最小空闲时间 -->

         <property name="minEvictableIdleTimeMillis" value="1800000" />

         <!-- 连接空闲多久后释放, 当空闲时间>该值 且 空闲连接>最大空闲连接数 时直接释放 -->

         <property name="softMinEvictableIdleTimeMillis" value="10000" />

         <!-- 获取连接时的最大等待毫秒数,小于零:阻塞不确定的时间,默认-1 -->

         <property name="maxWaitMillis" value="1500" />

         <!-- 在获取连接的时候检查有效性, 默认false -->

         <property name="testOnBorrow" value="true" />

         <!-- 在空闲时检查有效性, 默认false -->

         <property name="testWhileIdle" value="true" />

         <!-- 连接耗尽时是否阻塞, false报异常,ture阻塞直到超时, 默认true -->

         <property name="blockWhenExhausted" value="false" />

</bean>

<!-- redis集群 -->

<bean id="jedisCluster" class="redis.clients.jedis.JedisCluster">

         <constructor-arg index="0">

                  <set>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7001"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7002"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7003"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7004"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7005"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7006"></constructor-arg>

                          </bean>

                  </set>

         </constructor-arg>

         <constructor-arg index="1" ref="jedisPoolConfig"></constructor-arg>

</bean>


Ø 
测试代码

 

private ApplicationContext applicationContext;

         @Before

         publicvoid init() {

                  applicationContext = new ClassPathXmlApplicationContext(

                                   "classpath:applicationContext.xml");

         }

 

         // redis集群

         @Test

         publicvoid testJedisCluster() {

                  JedisCluster jedisCluster = (JedisCluster) applicationContext

                                   .getBean("jedisCluster");

 

                  jedisCluster.set("name", "zhangsan");

                  String value = jedisCluster.get("name");

                  System.out.println(value);

         }

 

 

1.1   redis-cluster投票:容错

 

最小节点数:3

(1)节点失效判断:集群中所有master参与投票,如果半数以上master节点与其中一个master节点通信超过(cluster-node-timeout),认为该master节点挂掉.

(2)集群失效判断:什么时候整个集群不可用(cluster_state:fail)? 

Ø  如果集群任意master挂掉,且当前master没有slave,则集群进入fail状态。也可以理解成集群的[0-16383]slot映射不完全时进入fail状态。

Ø  如果集群超过半数以上master挂掉,无论是否有slave,集群进入fail状态。

 

1.2   安装Ruby环境

redis集群需要使用集群管理脚本redis-trib.rb,它的执行相应依赖ruby环境。

 

l  第一步:安装ruby

yum install ruby

yum install rubygems

l  第三步:安装rubyredis的接口程序redis-3.2.2.gem

gem install redis -V 3.2.2

l  第四步:复制redis-3.2.9/src/redis-trib.rb文件到/usr/local/redis目录

cp redis-3.2.9/src/redis-trib.rb /usr/local/redis-cluster/ -r

 

1.3   安装Redis集群(RedisCluster

Redis集群最少需要三台主服务器,三台从服务器

端口号分别为:7001~7006

 

l  第一步:创建7001实例,并编辑redis.conf文件,修改port7001

注意:创建实例,即拷贝单机版安装时,生成的bin目录,为7001目录。

 

l  第二步:修改redis.conf配置文件,打开Cluster-enable yes

 

 

l  第三步:复制7001,创建7002~7006实例,注意端口修改

l  第四步:启动所有的实例

l  第五步:创建Redis集群

./redis-trib.rb create --replicas 1 192.168.10.133:7001 192.168.10.133:7002 192.168.10.133:7003 192.168.10.133:7004 192.168.10.133:7005  192.168.10.133:7006

>>> Creating cluster

Connecting to node 192.168.10.133:7001: OK

Connecting to node 192.168.10.133:7002: OK

Connecting to node 192.168.10.133:7003: OK

Connecting to node 192.168.10.133:7004: OK

Connecting to node 192.168.10.133:7005: OK

Connecting to node 192.168.10.133:7006: OK

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

Using 3 masters:

192.168.10.133:7001

192.168.10.133:7002

192.168.10.133:7003

Adding replica 192.168.10.133:7004 to 192.168.10.133:7001

Adding replica 192.168.10.133:7005 to 192.168.10.133:7002

Adding replica 192.168.10.133:7006 to 192.168.10.133:7003

M: d8f6a0e3192c905f0aad411946f3ef9305350420 192.168.10.133:7001

   slots:0-5460 (5461 slots) master

M: 7a12bc730ddc939c84a156f276c446c28acf798c 192.168.10.133:7002

   slots:5461-10922 (5462 slots) master

M: 93f73d2424a796657948c660928b71edd3db881f 192.168.10.133:7003

   slots:10923-16383 (5461 slots) master

S: f79802d3da6b58ef6f9f30c903db7b2f79664e61 192.168.10.133:7004

   replicates d8f6a0e3192c905f0aad411946f3ef9305350420

S: 0bc78702413eb88eb6d7982833a6e040c6af05be 192.168.10.133:7005

   replicates 7a12bc730ddc939c84a156f276c446c28acf798c

S: 4170a68ba6b7757e914056e2857bb84c5e10950e 192.168.10.133:7006

   replicates 93f73d2424a796657948c660928b71edd3db881f

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 192.168.10.133:7001)

M: d8f6a0e3192c905f0aad411946f3ef9305350420 192.168.10.133:7001

   slots:0-5460 (5461 slots) master

M: 7a12bc730ddc939c84a156f276c446c28acf798c 192.168.10.133:7002

   slots:5461-10922 (5462 slots) master

M: 93f73d2424a796657948c660928b71edd3db881f 192.168.10.133:7003

   slots:10923-16383 (5461 slots) master

M: f79802d3da6b58ef6f9f30c903db7b2f79664e61 192.168.10.133:7004

   slots: (0 slots) master

   replicates d8f6a0e3192c905f0aad411946f3ef9305350420

M: 0bc78702413eb88eb6d7982833a6e040c6af05be 192.168.10.133:7005

   slots: (0 slots) master

   replicates 7a12bc730ddc939c84a156f276c446c28acf798c

M: 4170a68ba6b7757e914056e2857bb84c5e10950e 192.168.10.133:7006

   slots: (0 slots) master

   replicates 93f73d2424a796657948c660928b71edd3db881f

[OK] All nodes agree about slots configuration.

>>> Check for open slots...

>>> Check slots coverage...

[OK] All 16384 slots covered.

[root@localhost-0723 redis]#

 

1.4   命令客户端连接集群

命令:./redis-cli –h 127.0.0.1 –p 7001 –c

注意:-c 表示是以redis集群方式进行连接

./redis-cli -p 7006 -c

127.0.0.1:7006> set key1 123

-> Redirected to slot [9189] located at 127.0.0.1:7002

OK

127.0.0.1:7002>

 

1.5   查看集群的命令

Ø  查看集群状态

127.0.0.1:7003> 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:3

cluster_stats_messages_sent:926

cluster_stats_messages_received:926

 

Ø  查看集群中的节点:

 

127.0.0.1:7003> cluster nodes

7a12bc730ddc939c84a156f276c446c28acf798c 127.0.0.1:7002 master - 0 1443601739754 2 connected 5461-10922

93f73d2424a796657948c660928b71edd3db881f 127.0.0.1:7003 myself,master - 0 0 3 connected 10923-16383

d8f6a0e3192c905f0aad411946f3ef9305350420 127.0.0.1:7001 master - 0 1443601741267 1 connected 0-5460

4170a68ba6b7757e914056e2857bb84c5e10950e 127.0.0.1:7006 slave 93f73d2424a796657948c660928b71edd3db881f 0 1443601739250 6 connected

f79802d3da6b58ef6f9f30c903db7b2f79664e61 127.0.0.1:7004 slave d8f6a0e3192c905f0aad411946f3ef9305350420 0 1443601742277 4 connected

0bc78702413eb88eb6d7982833a6e040c6af05be 127.0.0.1:7005 slave 7a12bc730ddc939c84a156f276c446c28acf798c 0 1443601740259 5 connected

127.0.0.1:7003>

 

1.6   维护节点

集群创建成功后可以继续向集群中添加节点

1.6.1添加主节点

l  先创建7007节点

l  添加7007结点作为新节点

执行命令:./redis-trib.rb add-node 127.0.0.1:7007 127.0.0.1:7001

 

l  查看集群结点发现7007已添加到集群中

 

 

1.6.2hash槽重新分配(数据迁移)

添加完主节点需要对主节点进行hash槽分配,这样该主节才可以存储数据

 

l  查看集群中槽占用情况

 

redis集群有16384个槽,集群中的每个结点分配自已槽,通过查看集群结点可以看到槽占用情况。

 

l  给刚添加的7007结点分配槽

* 第一步:连接上集群(连接集群中任意一个可用结点都行)

./redis-trib.rb reshard 192.168.10.133:7001

 

* 第二步:输入要分配的槽数量

 

输入:3000,表示要给目标节点分配3000个槽

 

* 第三步:输入接收槽的结点id

 

输入:15b809eadae88955e36bcdbb8144f61bbbaf38fb

PS:这里准备给7007分配槽,通过cluster nodes查看7007结点id为:

15b809eadae88955e36bcdbb8144f61bbbaf38fb

 

* 第四步:输入源结点id

 

输入:all

 

* 第五步:输入yes开始移动槽到目标结点id

 

输入:yes

 

1.6.3添加从节点

 

l  添加7008从结点,将7008作为7007的从结点

 

命令:./redis-trib.rb add-node --slave --master-id  主节点id   新节点的ip和端口   旧节点ip和端口(集群中任一节点都可以)

 

执行如下命令:

./redis-trib.rb add-node --slave --master-id  35da64607a02c9159334a19164e68dd95a3b943c 192.168.10.103:7008 192.168.10.103:7001

 

35da64607a02c9159334a19164e68dd95a3b943c7007结点的id,可通过cluster nodes查看。

 

 

 

注意:如果原来该结点在集群中的配置信息已经生成到cluster-config-file指定的配置文件中(如果cluster-config-file没有指定则默认为nodes.conf),这时可能会报错:

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

解决方法是删除生成的配置文件nodes.conf,删除后再执行./redis-trib.rb add-node指令

 

l  查看集群中的结点,刚添加的70087007的从节点:

 

 

 

1.6.4删除结点

命令:./redis-trib.rb del-node 127.0.0.1:7005 4b45eb75c8b428fbd77ab979b85080146a9bc017

 

 

删除已经占有hash槽的结点会失败,报错如下:

[ERR] Node 127.0.0.1:7005 is not empty! Reshard data away and try again.

 

需要将该结点占用的hash槽分配出去(参考hash槽重新分配章节)。

1.7   Jedis连接集群

需要开启防火墙,或者直接关闭防火墙。

service iptables stop

 

1.7.1代码实现

创建JedisCluster类连接redis集群。

 

@Test

publicvoid testJedisCluster() throws Exception {

         //创建一连接,JedisCluster对象,在系统中是单例存在

         Set<HostAndPort> nodes = new HashSet<>();

         nodes.add(new HostAndPort("192.168.10.133", 7001));

         nodes.add(new HostAndPort("192.168.10.133", 7002));

         nodes.add(new HostAndPort("192.168.10.133", 7003));

         nodes.add(new HostAndPort("192.168.10.133", 7004));

         nodes.add(new HostAndPort("192.168.10.133", 7005));

         nodes.add(new HostAndPort("192.168.10.133", 7006));

         JedisCluster cluster = new JedisCluster(nodes);

         //执行JedisCluster对象中的方法,方法和redis一一对应。

         cluster.set("cluster-test", "my jedis cluster test");

         String result = cluster.get("cluster-test");

         System.out.println(result);

         //程序结束时需要关闭JedisCluster对象

         cluster.close();

}

1.7.2使用spring

Ø  配置applicationContext.xml

<!-- 连接池配置 -->

<bean id="jedisPoolConfig" class="redis.clients.jedis.JedisPoolConfig">

         <!-- 最大连接数 -->

         <property name="maxTotal" value="30" />

         <!-- 最大空闲连接数 -->

         <property name="maxIdle" value="10" />

         <!-- 每次释放连接的最大数目 -->

         <property name="numTestsPerEvictionRun" value="1024" />

         <!-- 释放连接的扫描间隔(毫秒) -->

         <property name="timeBetweenEvictionRunsMillis" value="30000" />

         <!-- 连接最小空闲时间 -->

         <property name="minEvictableIdleTimeMillis" value="1800000" />

         <!-- 连接空闲多久后释放, 当空闲时间>该值 且 空闲连接>最大空闲连接数 时直接释放 -->

         <property name="softMinEvictableIdleTimeMillis" value="10000" />

         <!-- 获取连接时的最大等待毫秒数,小于零:阻塞不确定的时间,默认-1 -->

         <property name="maxWaitMillis" value="1500" />

         <!-- 在获取连接的时候检查有效性, 默认false -->

         <property name="testOnBorrow" value="true" />

         <!-- 在空闲时检查有效性, 默认false -->

         <property name="testWhileIdle" value="true" />

         <!-- 连接耗尽时是否阻塞, false报异常,ture阻塞直到超时, 默认true -->

         <property name="blockWhenExhausted" value="false" />

</bean>

<!-- redis集群 -->

<bean id="jedisCluster" class="redis.clients.jedis.JedisCluster">

         <constructor-arg index="0">

                  <set>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7001"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7002"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7003"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7004"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7005"></constructor-arg>

                          </bean>

                          <bean class="redis.clients.jedis.HostAndPort">

                                   <constructor-arg index="0" value="192.168.101.3"></constructor-arg>

                                   <constructor-arg index="1" value="7006"></constructor-arg>

                          </bean>

                  </set>

         </constructor-arg>

         <constructor-arg index="1" ref="jedisPoolConfig"></constructor-arg>

</bean>

 

 

Ø  测试代码

 

private ApplicationContext applicationContext;

         @Before

         publicvoid init() {

                  applicationContext = new ClassPathXmlApplicationContext(

                                   "classpath:applicationContext.xml");

         }

 

         // redis集群

         @Test

         publicvoid testJedisCluster() {

                  JedisCluster jedisCluster = (JedisCluster) applicationContext

                                   .getBean("jedisCluster");

 

                  jedisCluster.set("name", "zhangsan");

                  String value = jedisCluster.get("name");

                  System.out.println(value);

         }

 

posted @ 2020-01-31 22:31  G丶影  阅读(341)  评论(0编辑  收藏  举报