Redis集群的操作

1、原始集群

      1.1    启动Redis集群

[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7001/redis.conf 
[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7002/redis.conf 
[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7003/redis.conf 
[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7004/redis.conf 
[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7005/redis.conf 
[root@cent03 ~]#  /usr/local/redis/bin/redis-server /usr/local/redis-cluster/7006/redis.conf

      1.2    查看server执行状态

[root@cent03 ~]# ps -ef | grep redis
root      2322     1  0 04:48 ?        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7001 [cluster]            
root      2326     1  0 04:48 ?
        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7002 [cluster]            
root      2330     1  0 04:48 ?        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7003 [cluster]            
root      2334     1  0 04:48 ?        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7004 [cluster]            
root      2339     1  0 04:48 ?        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7005 [cluster]            
root      2344     1  0 04:48 ?        00:00:00 /usr/local/redis/bin/redis-server 192.168.2.105:7006 [cluster]            
root      2349  2299  0 04:49 pts/0    00:00:00 grep redis

      1.3    查看集群状态

# -c 表示集群模式

[root@cent03 ~]# /usr/local/redis/bin/redis-cli -c -h 192.168.2.105 -p 7001
192.168.2.105:7001> cluster nodes
aae7ff08b8cc4bb5473044a52f926b6e82c96840 192.168.2.105:7001 myself,slave 1b362004cf174d9ecb72c0650117781618c52a70 0 0 1 connected
1b362004cf174d9ecb72c0650117781618c52a70 192.168.2.105:7004 master - 0 1477568990462 8 connected 0-5460
987b27c96774f9c9bbbbf7bf3ccfe4e4b2e4c3e7 192.168.2.105:7005 slave f9a0fbd03254b0010bca6782867dfa65f21a2448 0 1477568991974 5 connected
f9a0fbd03254b0010bca6782867dfa65f21a2448 192.168.2.105:7002 master - 0 1477568991471 2 connected 5461-10922
cabff9be14080ae12bfad0b199ae737639ec57bd 192.168.2.105:7003 master - 0 1477568990967 3 connected 10923-16383
286b1ed602c9e74f23b8bd8d66db50308b256d3e 192.168.2.105:7006 slave cabff9be14080ae12bfad0b199ae737639ec57bd 0 1477568990967 6 connected
192.168.2.105:7001> 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:8
cluster_my_epoch:8
cluster_stats_messages_sent:998
cluster_stats_messages_received:351

2、新集群操作

     2.1  我们新建俩个服务，依照之前搭建的集群方式新增俩个节点：（一主一从 master、slave）  Master：7007      Slave：7008     

             步骤一：创建7007/7008文件夹。

拷贝redis.conf文件到对于的7007,7008文件夹下要进行改动配置文件。

             [root@cent03 redis-cluster]# mkdir 7007
             [root@cent03 redis-cluster]# mkdir 7008
             [root@cent03 redis-cluster]# cd 7001
             [root@cent03 7001]# cp redis.conf /usr/local/redis-cluster/7007/
             [root@cent03 7001]# cp redis.conf /usr/local/redis-cluster/7008/
             [root@cent03 7001]# vim/usr/local/redis-cluster/7007/redis.conf
                改动内容例如以下：
                port:7007
                dir /usr/local/redis-cluster/7007/
                cluster-config-file nodes-7007.conf
             [root@cent03 7001]# vim/usr/local/redis-cluster/7008/redis.conf
                改动内容例如以下：
                port:7008
                dir /usr/local/redis-cluster/7008/
                cluster-config-file nodes-7008.conf

           步骤二：启动7007和7008俩个服务并查看服务状态。

[root@cent03 7001]#/usr/local/redis/bin/redis-server /usr/local/redis-cluster/7007/redis.conf
[root@cent03 7001]#/usr/local/redis/bin/redis-server /usr/local/redis-cluster/7008/redis.conf
[root@cent03 7001]# ps -el | grep redis

     2.2  学习redis-trib命令使用

[root@cent03 ~]# cd /usr/local/redis-3.0.0-rc2/src/
[root@cent03 src]# ./redis-trib.rb 
Usage: redis-trib <command> <options> <arguments ...>

  call            host:port command arg arg .. arg
  set-timeout     host:port milliseconds
  del-node        host:port node_id
  fix             host:port
  reshard         host:port
                  --slots <arg>
                  --yes
                  --from <arg>
                  --to <arg>
  help            (show this help)
  add-node        new_host:new_port existing_host:existing_port
                  --slave
                  --master-id <arg>
  check           host:port
  import          host:port
                  --from <arg>
  create          host1:port1 ... hostN:portN
                  --replicas <arg>

For check, fix, reshard, del-node, set-timeout you can specify the host and port of any working node in the cluster.

             1. create：创建一个集群环境host1:port1 ... hostN:portN（集群中的主从节点比例）

             2. call：能够运行redis命令

             3. add-node：将一个节点加入到集群里。第一个參数为新节点的ip:port，第二个參数为集群中随意一个已经存在的节点的ip:port

             4. del-node：移除一个节点

             5. reshard：又一次分片

             6. check：检查集群状态

3、新增一个主节点7007（master）

           步骤一：使用add-node命令：绿色为新增节点。红色为已知存在节点

            /usr/local/redis-3.0.0-rc2/src/redis-trib.rb  add-node 192.168.2.105:7007 192.168.2.105:7001

>>> Adding node 192.168.2.105:7007 to cluster 192.168.2.105:7004
Connecting to node 192.168.2.105:7004: OK
Connecting to node 192.168.2.105:7005: OK
Connecting to node 192.168.2.105:7002: OK
Connecting to node 192.168.2.105:7006: OK
Connecting to node 192.168.2.105:7001: OK
Connecting to node 192.168.2.105:7003: OK
>>> Performing Cluster Check (using node 192.168.2.105:7004)
M: 1b362004cf174d9ecb72c0650117781618c52a70 192.168.2.105:7004
   slots:0-5460 (5461 slots) master
   1 additional replica(s)
S: 987b27c96774f9c9bbbbf7bf3ccfe4e4b2e4c3e7 192.168.2.105:7005
   slots: (0 slots) slave
   replicates f9a0fbd03254b0010bca6782867dfa65f21a2448
M: f9a0fbd03254b0010bca6782867dfa65f21a2448 192.168.2.105:7002
   slots:5461-10922 (5462 slots) master
   1 additional replica(s)
S: 286b1ed602c9e74f23b8bd8d66db50308b256d3e 192.168.2.105:7006
   slots: (0 slots) slave
   replicates cabff9be14080ae12bfad0b199ae737639ec57bd
S: aae7ff08b8cc4bb5473044a52f926b6e82c96840 192.168.2.105:7001
   slots: (0 slots) slave
   replicates 1b362004cf174d9ecb72c0650117781618c52a70
M: cabff9be14080ae12bfad0b199ae737639ec57bd 192.168.2.105:7003
   slots:10923-16383 (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.
Connecting to node 192.168.2.105:7007: OK
>>> Send CLUSTER MEET to node 192.168.2.105:7007 to make it join the cluster.
[OK] New node added correctly.

           步骤二：查看集群状态：

/usr/local/redis/bin/redis-cli -c -h 192.168.2.105 -p 7001
192.168.2.105:7001> cluster  nodes
192.168.2.105:7001> cluster nodes
aae7ff08b8cc4bb5473044a52f926b6e82c96840 192.168.2.105:7001 myself,slave 1b362004cf174d9ecb72c0650117781618c52a70 0 0 1 connected
c25aca95364e13b58e57d831126ed77656ba1d76 192.168.2.105:7007 master - 0 1477571648229 9 connected 0-65 5461-5527 10923-10988
1b362004cf174d9ecb72c0650117781618c52a70 192.168.2.105:7004 master - 0 1477571648734 8 connected 66-5460
987b27c96774f9c9bbbbf7bf3ccfe4e4b2e4c3e7 192.168.2.105:7005 slave f9a0fbd03254b0010bca6782867dfa65f21a2448 0 1477571649238 5 connected
f9a0fbd03254b0010bca6782867dfa65f21a2448 192.168.2.105:7002 master - 0 1477571649238 2 connected 5528-10922
cabff9be14080ae12bfad0b199ae737639ec57bd 192.168.2.105:7003 master - 0 1477571648229 3 connected 10989-16383
286b1ed602c9e74f23b8bd8d66db50308b256d3e 192.168.2.105:7006 slave cabff9be14080ae12bfad0b199ae737639ec57bd 0 1477571647225 6 connected

            注意：当加入节点成功以后，新增的节点不会有不论什么数据。由于它没有分配不论什么的slot（hash槽）。我们须要为新节点手工分配slot。

4、为7007分配slot槽

     步骤一：使用redis-trib 命令。找到集群中的随意一个主节点（红色位置表现集群中的随意一个主节点），对其进行又一次分片工作。

     [root@cent03 7001]#/usr/local/redis-3.0.0-rc2/src/redis-trib.rb reshard 192.168.2.105:7004

Connecting to node 192.168.2.105:7004: OK
Connecting to node 192.168.2.105:7007: OK
Connecting to node 192.168.2.105:7005: OK
Connecting to node 192.168.2.105:7002: OK
Connecting to node 192.168.2.105:7006: OK
Connecting to node 192.168.2.105:7001: OK
Connecting to node 192.168.2.105:7003: OK
>>> Performing Cluster Check (using node 192.168.2.105:7004)
M: 1b362004cf174d9ecb72c0650117781618c52a70 192.168.2.105:7004
   slots:0-5460 (5461 slots) master
   1 additional replica(s)
M: c25aca95364e13b58e57d831126ed77656ba1d76 192.168.2.105:7007
   slots: (0 slots) master
   0 additional replica(s)
S: 987b27c96774f9c9bbbbf7bf3ccfe4e4b2e4c3e7 192.168.2.105:7005
   slots: (0 slots) slave
   replicates f9a0fbd03254b0010bca6782867dfa65f21a2448
M: f9a0fbd03254b0010bca6782867dfa65f21a2448 192.168.2.105:7002
   slots:5461-10922 (5462 slots) master
   1 additional replica(s)
S: 286b1ed602c9e74f23b8bd8d66db50308b256d3e 192.168.2.105:7006
   slots: (0 slots) slave
   replicates cabff9be14080ae12bfad0b199ae737639ec57bd
S: aae7ff08b8cc4bb5473044a52f926b6e82c96840 192.168.2.105:7001
   slots: (0 slots) slave
   replicates 1b362004cf174d9ecb72c0650117781618c52a70
M: cabff9be14080ae12bfad0b199ae737639ec57bd 192.168.2.105:7003
   slots:10923-16383 (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.

（提示一）

How many slots do youwant to move (from 1 to 16384)?200

（提示二）

What is the receivingnode ID? 382634a4025778c040b7213453fd42a709f79e28

Please enter all thesource node IDs.

  Type ‘all‘ to use all the nodes as sourcenodes for the hash slots.

  Type ‘done‘ once you entered all the sourcenodes IDs.

Source node #1:all

Ready to move 200slots.

  Sourcenodes:

    M: 614d0def75663f2620b6402a017014b57c912dad192.168.2.105:7001

   slots:0-5460 (5461 slots) master

   1 additional replica(s)

    M: 8aac82b63d42a1989528cd3906579863a5774e77192.168.2.105:7002

   slots:5461-10922 (5462 slots) master

   1 additional replica(s)

    M: 83df08875c7707878756364039df0a4c8658f272192.168.2.105:7003

   slots:10923-16383 (5461 slots) master

   1 additional replica(s)

  Destinationnode:

    M: 382634a4025778c040b7213453fd42a709f79e28192.168.2.105:7007

   slots: (0 slots) master

   0 additional replica(s)

  Resharding plan:（分片运行计划日志）

Moving slot 5461 from8aac82b63d42a1989528cd3906579863a5774e77

...

Moving slot 0 from 614d0def75663f2620b6402a017014b57c912dad

...

    Moving slot 10923 from83df08875c7707878756364039df0a4c8658f272

...

（提示三）

Do you want toproceed with the proposed reshard plan (yes/no)?yes

Moving slot 65 from 192.168.2.105:7001to 192.168.2.105:7007:

...

Moving slot 10923 from 192.168.2.105:7003to 192.168.2.105:7007:

...

Moving slot 5527 from 192.168.2.105:7002to 192.168.2.105:7007:

提示一：是希望你须要多少个槽移动到新的节点上，能够自己设置，比方200个槽。

提示二：是你须要把这200个slot槽移动到那个节点上去（须要指定节点id），而且下个提示是输入all为从全部主节点（7001 7002 7003）中分别抽取响应的槽数（一共为200个槽到指定的新节点中。，而且会打印运行分片的计划。

）

      如上图所看到的。如今我们的7007已经有slot槽了，也就是说能够在7007上进行读写数据啦。到此为止我们的7007已经增加到集群中啦，而且是主节点（Master）

5、加入从节点（7008）到集群中去

          步骤一：还是须要运行add-node命令：

[root@cent03 7001]#     /usr/local/redis-3.0.0-rc2/src/redis-trib.rbadd-node 192.168.2.105:7008 192.168.2.105:7004

         步骤二：我们须要运行replicate命令来指定当前节点（从节点）的主节点id为哪个。

       首先须要登录新加的7008节点的client。然后使用集群命令进行操作。把当前的7008（slave）节点指定到一个主节点下（这里使用之前创建的7007主节点，红色表示节点id）

[root@cent03 ~]# /usr/local/redis/bin/redis-cli -c -h 192.168.2.105 -p 7008
192.168.2.105:7008> cluster replicate <strong><span style="color:#ff0000;">c25aca95364e13b58e57d831126ed77656ba1d76</span></strong>
192.168.2.105:7008> OK（提示OK则操作成功）

我们继续看一下当前集群的状态，例如以下图：我们已经成功的把7008放到7007这个主节点以下了，到此为止我们已经成功的加入完一个从节点了。

6、尝试删除一个节点（7008 slave）

          步骤一：删除从节点7008，输入del-node命令，指定删除节点ip和port，以及节点id（红色为7008节点id）

[root@cent03 7001]# /usr/local/redis-3.0.0-rc2/src/redis-trib.rb  del-node 192.168.2.105:7008bd83784a3eb65264fcdfc38e6c2a8f6f5dc6dd0c

          步骤二：再次查看一下集群状态，例如以下图所看到的，我们已经成功的移除了7008 slave节点。另外我们发现移除一个节点以后，当前节点的服务进程也会随之销毁。

能够使用ps命令查看当前的服务（ps -ef | grep redis），发现少了一个执行的server，也就是刚移除的7008从节点。

7、删除之前增加的主节点7007。这个步骤会相对照较麻烦一些，由于主节点的里面是有分配了slot槽的，所以我们这里必须先把7007里的slot槽放入到其它的可用主节点中去。然后再进行移除节点操作才行，不然会出现数据丢失问题。

         步骤一：删除7007（master）节点之前，我们须要先把其所有的数据（slot槽）移动到其它节点上去（眼下仅仅能把master的数据迁移到一个节点上，临时做不了平均分配功能）。

[root@cent03 7001]# /usr/local/redis-3.0.0-rc2/src/redis-trib.rb  reshard 192.168.2.105:7007

输出例如以下：

>>> Performing Cluster Check(using node 192.168.2.105:7007)

M: 382634a4025778c040b7213453fd42a709f79e28192.168.2.105:7007

   slots:0-65,5461-5527,10923-10988 (199 slots) master

   0 additional replica(s)

S: fa299e41c173fa807ba04684c2f5e5e185d5f7d0192.168.2.105:7006

  slots: (0 slots) slave

  replicates 83df08875c7707878756364039df0a4c8658f272

S: a69b98937844c6050ee5885266ccccb185a3f36a192.168.2.105:7004

  slots: (0 slots) slave

  replicates 614d0def75663f2620b6402a017014b57c912dad

M: 614d0def75663f2620b6402a017014b57c912dad192.168.2.105:7001

  slots:66-5460 (5395 slots) master

   1additional replica(s)

M: 8aac82b63d42a1989528cd3906579863a5774e77192.168.2.105:7002

  slots:5528-10922 (5395 slots) master

   1additional replica(s)

S: adb99506ddccad332e09258565f2e5f4f456a150192.168.2.105:7005

  slots: (0 slots) slave

   replicates8aac82b63d42a1989528cd3906579863a5774e77

M: 83df08875c7707878756364039df0a4c8658f272192.168.2.105:7003

  slots:10989-16383 (5395 slots) master

   1additional replica(s)

[OK] All nodes agree about slotsconfiguration.

>>> Check for open slots...

>>> Check slots coverage...

[OK] All 16384 slots covered.

How many slots do youwant to move (from 1 to 16384)?199

（凝视：这里不会是正好200个槽）

What is the receivingnode ID? 614d0def75663f2620b6402a017014b57c912dad
（凝视：这里是须要把数据移动到哪？7001的主节点id）

Please enter all thesource node IDs.

  Type ‘all‘ to use all the nodes as sourcenodes for the hash slots.

  Type ‘done‘ once you entered all the sourcenodes IDs.

Source node #1:382634a4025778c040b7213453fd42a709f79e28

（凝视：这里是须要数据源，也就是我们的7007节点id）

Source node #2:done

（凝视：这里直接输入done 開始生成迁移计划）

Ready to move 199 slots.

  Source nodes:

    M: 382634a4025778c040b7213453fd42a709f79e28192.168.2.105:7007

   slots:0-65,5461-5527,10923-10988 (199 slots)master

   0 additional replica(s)

  Destinationnode:

    M: 614d0def75663f2620b6402a017014b57c912dad192.168.2.105:7001

   slots:66-5460 (5395 slots) master

   1 additional replica(s)

 Resharding plan:

Moving slot 0 from 382634a4025778c040b7213453fd42a709f79e28

...

Do you want toproceed with the proposed reshard plan (yes/no)? Yes

（凝视：这里输入yes開始迁移）

Moving slot 0 from 192.168.2.105:7007 to 192.168.2.105:7001: 

         到此为止我们已经成功的把7007主节点的数据迁移到7001上去了，我们能够看一下如今的集群状态例如以下图。你会发现7007以下已经没有不论什么数据（slot）槽了，证明迁移成功！

          步骤二：最后我们直接使用del-node命令删除7007主节点就可以（红色表示7007的节点id）。

[root@cent03 7001]# /usr/local/redis-3.0.0-rc2/src/redis-trib.rbdel-node   192.168.2.105:7007   c25aca95364e13b58e57d831126ed77656ba1d76

最后：我们查看集群状态，一切还原为最初始状态啦！OK结束！

192.168.2.105:7001> cluster nodes
aae7ff08b8cc4bb5473044a52f926b6e82c96840 192.168.2.105:7001 myself,slave 1b362004cf174d9ecb72c0650117781618c52a70 0 0 1 connected
1b362004cf174d9ecb72c0650117781618c52a70 192.168.2.105:7004 master - 0 1477574617571 10 connected 0-5527 10923-10988
987b27c96774f9c9bbbbf7bf3ccfe4e4b2e4c3e7 192.168.2.105:7005 slave f9a0fbd03254b0010bca6782867dfa65f21a2448 0 1477574618076 5 connected
f9a0fbd03254b0010bca6782867dfa65f21a2448 192.168.2.105:7002 master - 0 1477574618579 2 connected 5528-10922
cabff9be14080ae12bfad0b199ae737639ec57bd 192.168.2.105:7003 master - 0 1477574617066 3 connected 10989-16383
286b1ed602c9e74f23b8bd8d66db50308b256d3e 192.168.2.105:7006 slave cabff9be14080ae12bfad0b199ae737639ec57bd 0 1477574617066 6 connected