Mysql Cluster水平扩展-添加数据节点

原文地址

http://blog.csdn.net/yabingshi_tech/article/details/47953109

一：实验环境

 

 

已搭建好MySQL cluster（一个管理节点，2个sql节点及2个数据节点）,搭建步骤请参考：http://blog.csdn.net/yabingshi_tech/article/details/47952135，现要在此基础上添加

192.168.6.70,192.168.6.71作为数据节点。(由于副本集参数已经配置成了2，所以必须一次添加2个数据节点）

二：实验步骤

2.1：解压cluster安装包

tar -xvf mysql-cluster-gpl-7.2.15-linux2.6-x86_64.tar.gz 

mv mysql-cluster-gpl-7.2.15-linux2.6-x86_64 /usr/local/mysql

 

2.2：修改管理节点配置文件

cd /data/mysql/mysql-cluster

 

修改全局配置文件vi mgmd.cnf，修改后的配置文件如下：

 [ndbd default]

 NoOfReplicas=2

 DataMemory=80M

 IndexMemory=18M

 [ndb_mgmd]

 hostname=192.168.6.66
 datadir=/data/mysql/mysql-cluster

 [ndbd]

 hostname=192.168.6.68

 datadir=/usr/local/mysql/data

 [ndbd]

 hostname=192.168.6.69

 datadir=/usr/local/mysql/data

 [ndbd]

 hostname=192.168.6.70

 datadir=/usr/local/mysql/data

 [ndbd]

 hostname=192.168.6.71

 datadir=/usr/local/mysql/data

 [mysqld]

 hostname=192.168.6.66

 [mysqld]

 hostname=192.168.6.67

 

 

2.3：重新启动管理节点服务

 

2.3.1 停止管理节点

 

ndb_mgm> 1 stop  

Node 1 has shutdown.  

Disconnecting to allow Management Server to shutdown  

 

--停止管理节点不影响sql节点的正常访问。

 

2.3.2 重新加载配置文件

[root@ser6-66 mysql-cluster]# ndb_mgmd -f mgmd.cnf --initial

MySQL Cluster Management Server mysql-5.5.35 ndb-7.2.15

/*

用initial或者reload参数重新加载

如果只在配置文件中新增一个数据节点，会报错：

[root@ser6-66 mysql-cluster]#  ndb_mgmd -f mgmd.cnf --initial

MySQL Cluster Management Server mysql-5.5.35 ndb-7.2.15

2015-08-18 16:18:37 [MgmtSrvr] ERROR    -- at line 38: Nodegroup 1 has 1 members, NoOfReplicas=2

2015-08-18 16:18:37 [MgmtSrvr] ERROR    -- Could not load configuration from 'mgmd.cnf'

*/

 

2.3.4 查看集群状态

 

 [root@ser6-66 mysql-cluster]# ndb_mgm 192.168.6.66
 -- NDB Cluster -- Management Client --
 ndb_mgm> show
 Connected to Management Server at: 192.168.6.66:1186
 Cluster Configuration
 ---------------------
 [ndbd(NDB)]4 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 id=4 (not connected, accepting connect from 192.168.6.70)
 id=5 (not connected, accepting connect from 192.168.6.71)

 [ndb_mgmd(MGM)]1 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]2 node(s)
 id=6 (not connected, accepting connect from 192.168.6.66)
 id=7 (not connected, accepting connect from 192.168.6.67)

发现我的两个sql节点not connect了，网上别人的都没断啊。以保万一，这个还是晚上访问量比较少的时候添数据节点加比较好。

2.4：修改配置文件/etc/my.cnf

在192.168.6.70,192.168.6.71上

编辑/etc/my.cnf

内容如下：

 [mysqld]
 ndbcluster                          #运行NDB存储引擎
 ndb-connectstring=192.168.6.66    #定位管理节点
 [mysql_cluster]
 ndb-connectstring=192.168.6.66    #定位管理节点

2.5：轮流重新启动已连接的数据节点

 ndb_mgm> 2 stop
 Node 2: Node shutdown initiated
 Node 2: Node shutdown completed.
 Node 2 has shutdown.

 [root@ser6-68 ~]# /usr/local/mysql/bin/ndbd   --ndb-connectstring=192.168.6.66
 2015-08-18 16:56:31 [ndbd] INFO     -- Angel connected to '192.168.6.66:1186'
 2015-08-18 16:56:31 [ndbd] INFO     -- Angel allocated nodeid: 2
 ndb_mgm> show
 Cluster Configuration
 ---------------------
 [ndbd(NDB)]4 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 id=4 (not connected, accepting connect from 192.168.6.70)
 id=5 (not connected, accepting connect from 192.168.6.71)

 [ndb_mgmd(MGM)]1 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]3 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)
 id=7 (not connected, accepting connect from 192.168.6.67)
 id=8 (not connected, accepting connect from any host)
 --发现sql节点也起来了一个。
 ndb_mgm> 3 stop
 Node 3: Node shutdown initiated
 Node 3: Node shutdown completed.
 Node 3 has shutdown.
 [root@ser6-69 download]# /usr/local/mysql/bin/ndbd   --ndb-connectstring=192.168.6.66
 2015-08-18 16:57:54 [ndbd] INFO     -- Angel connected to '192.168.6.66:1186'
 2015-08-18 16:57:54 [ndbd] INFO     -- Angel allocated nodeid: 3

 ndb_mgm> show
 Cluster Configuration
 ---------------------
 [ndbd(NDB)] 4 node(s)
 id=2    @192.168.6.68  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 id=3    @192.168.6.69  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 id=4 (not connected, accepting connect from 192.168.6.70)
 id=5 (not connected, accepting connect from 192.168.6.71)

 [ndb_mgmd(MGM)] 1 node(s)
 id=1    @192.168.6.66  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]   3 node(s)
 id=6    @192.168.6.66  (mysql-5.5.35 ndb-7.2.15)
 id=7    @192.168.6.67  (mysql-5.5.35 ndb-7.2.15)
 id=8 (not connected, accepting connect from any host)

 此时发现id=7的sql节点也自动连接上了。

 

2.6：轮流重新启动连接的sql节点

 service mysql restart
 ndb_mgm> show
 Cluster Configuration
 ---------------------
 [ndbd(NDB)]4 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 id=4 (not connected, accepting connect from 192.168.6.70)
 id=5 (not connected, accepting connect from 192.168.6.71)

 [ndb_mgmd(MGM)]1 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]3 node(s)
 id=6 (not connected, accepting connect from 192.168.6.66)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 

2.7：初始化新的数据节点

 在192.168.6.70,192.168.6.71上分别：
  /usr/local/mysql/bin/ndbd  --initial --ndb-connectstring=192.168.6.66
 ndb_mgm> show
 Cluster Configuration
 ---------------------
 [ndbd(NDB)]4 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, no nodegroup)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, no nodegroup)

 [ndb_mgmd(MGM)]1 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]3 node(s)
 id=6 (not connected, accepting connect from 192.168.6.66)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)
 /*
 mycluster 中没有参数设置节点组数. 节点组数=节点数/NoOfReplicas.
 所以你如果总共有两个数据节点,如果NoOfReplicas设为2, 那么意味着只有一个节点组. 如果NoOfReplicas=1,那么就有两个节点组。
 */

 ndb_mgm> CREATE NODEGROUP 4,5
 Nodegroup 1 created

 这个数字不是随意设置的，是指你要添加哪些node到nodegroup中，这里是id=4,id=5的数据节点。

 ndb_mgm> show
 Cluster Configuration
 ---------------------
 [ndbd(NDB)]4 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0, *)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 0)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 1)
 [email protected]  (mysql-5.5.35 ndb-7.2.15, Nodegroup: 1)

 [ndb_mgmd(MGM)]1 node(s)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 [mysqld(API)]3 node(s)
 id=6 (not connected, accepting connect from 192.168.6.66)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)
 [email protected]  (mysql-5.5.35 ndb-7.2.15)

 mysql> ALTER ONLINE TABLE t2 REORGANIZE PARTITION;
 Query OK, 0 rows affected (7.99 sec)
 Records: 0  Duplicates: 0  Warnings: 0

 mysql> select partition_name,table_rows from information_schema.PARTITIONS where     table_name='t2' and table_schema='dba';
 +----------------+------------+
 | partition_name | table_rows |
 +----------------+------------+
 | p0             |          3 |
 | p1             |          0 |
 | p2             |          1 |
 | p3             |          2 |
 +----------------+------------+
 4 rows in set (0.02 sec)

原先t2分布在p0,p1,现在也能分布在p2,p3啦。成功啦！嘎嘎！

 

注意:如果配置过程中出现问题，记得将防火墙给关闭。

 /*

以下是我在配置过程中遇到的问题：

ndb_mgm>  create nodegroup 6,7

*   110: Error

*        Time out talking to management server

mysql> ALTER ONLINE TABLE t2 REORGANIZE PARTITION;

ERROR 1297 (HY000): Got temporary error 711 'System busy with node restart, schema operations not allowed' from NDBCLUSTER

出错原因:192.168.6.70,192.168.6.71的防火墙处于开启状态

解决办法：关闭防火墙

*/

本篇文章参考：

http://blog.csdn.net/chengfei112233/article/details/7706275