1. 先了解一下你是否应该用MySQL集群。
减少数据中心结点压力和大数据量处理,采用把MySQL分布,一个或多个application对应一个MySQL数据库。把几个MySQL数据库公用的数据做出共享数据,例如购物车,用户对象等等,存在数据结点里面。其他不共享的数据还维持在各自分布的MySQL数据库本身中。
2. 集群MySQL中名称概念.(如上图)
1)Sql结点(SQL node--上图对应为MySQLd):分布式数据库。包括自身数据和查询中心结点数据.
2)数据结点(Data node -- ndbd):集群共享数据(内存中).
3)管理服务器(Management Server – ndb_mgmd):集群管理SQL node,Data node.
3.配置
MySQL-max版本,当然现在MySQL集群系统windonws平台上面不被支持.
安装MySQL就不多说了,网上一大堆,简明扼要。
- A:192.168.1.251 – Data node和Management Server.
- B:192.168.1.254 – SQL node.
当然,你也可以让一个机器同时为3者。
A,B my.inf加上:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
|
[MySQLD]
ndbcluster # run NDB engine
ndb-connectstring=192.168.1.251 # location of MGM node
# Options for ndbd process:
[MySQL_CLUSTER]
ndb-connectstring=192.168.1.251 # location of MGM node
A: /var/lib/MySQL-cluster/config.ini
[NDBD DEFAULT]
NoOfReplicas=1 # Number of replicas
DataMemory=80M # How much memory to allocate for data storage
IndexMemory=18M # How much memory to allocate for index storage
# For DataMemory and IndexMemory, we have used the
# default values. Since the "world" database takes up
# only about 500KB, this should be more than enough for
# this example Cluster setup.
# TCP/IP options:
[TCP DEFAULT]
portnumber=2202 # This the default; however, you can use any
# port that is free for all the hosts in cluster
# Note: It is recommended beginning with MySQL 5.0 that
# you do not specify the portnumber at all and simply allow
# the default value to be used instead
# Management process options:
[NDB_MGMD]
hostname=192.168.1.251 # Hostname or IP address of MGM node
datadir=/var/lib/MySQL-cluster # Directory for MGM node logfiles
# Options for data node "A":
[NDBD]
# (one [NDBD] section per data node)
hostname=192.168.1.251 # Hostname or IP address
datadir=/usr/local/MySQL/data # Directory for this data node's datafiles
# SQL node options:
[MySQLD]
hostname=192.168.1.254
#[MySQLD] #这个相当于192.168.1.251
|
4. 启动测试
在管理服务器上面(这里是192.168.1.251):
1
|
shell>ndb_mgmd -f /var/lib/MySQL-cluster/config .ini
|
在数据结点服务器上面(依然是192.168.1.251and more):
1
|
shell>ndbd--initial
|
(第一次时加--initial参数)
SQL结点服务器上面(192.168.1.254):
1
|
shell>MySQLd &
|
在251上面察看
1
|
. /ndb_mgm
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
-- NDB Cluster -- Management Client --
ndb_mgm> show
Connected to Management Server at: 192.168.1.251:1186
Cluster Configuration
---------------------
[ndbd(NDB)] 1 node(s)
id=2 @192.168.1.251 (Version:5.0.22, Nodegroup: 0, Master)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.1.251 (Version:5.0.22)
[MySQLd(API)] 1 node(s)
id=3 @192.168.1.254 (Version:5.0.22)
ok
|
关闭集群:
1
|
shell>ndb_mgm -e shutdown
|
5.基本的集群说明
1)在MySQL集群中.当table引擎为NDBCLUSTER时才做集群,其他非NDBCLUSTER表和一般MySQL数据库表一样,不会共享数据. NDBCLUSTER表数据存储在Data node服务器内存中,Data Node可以为1台或多台服务器,它们之间存放共享数据。Data Node服务器可以分组数据copy。
例如:2,3,4,5为四台Data Node服务器ID. 2,3为组0。 4,5为组1。 2,3维持数据相同,4,5维持数据相同。 组0和组1维持数据不同。
2)sql node服务器中,非NDBCLUSTER数据存在本身数据库中,table引擎为NDBCLUSTER时,数据存储在Data Node中。当查询NDBCLUSTER表时,它会从Data node集群中提起数据.
3)Manager server
管理SQl node和Data node状态。
附:MySQL集群配置详细过程录制
1、准备三台linux服务器(三台机器进行如下配置)
--hostname配置
192.168.9.241 sqltest01 (mysqld及存储节点)
192.168.9.242 sqltest02 (mysqld及存储节点)
192.168.9.243 sqltest03
其中,sqltest01、sqltest02分别是mysql节点及存储节点,sqltest03为管理节点
--同时,把防火墙进行关闭或者把相关的端口打开,如3306,管理节点的1186等
1
|
[root@sqltest01 u01] # service iptables status
|
1
|
iptables: Firewall is not running.
|
如果开启的,请使用service iptables stop
--创建相应的用户及目录
1
2
3
4
|
[root@sqltest01 u01] # groupadd mysql
[root@sqltest01 u01] # useradd -r -g mysql mysql
[root@sqltest01 u01] # mkdir -p /usr/local/mysql
[root@sqltest01 u01] # chown -R mysql.mysql
|
2、mysql cluster
下载网址:dev.mysql.com,然后选择cluster,然后在网页中出现的选择平台中,选择linux generic!在这里选择所需要tar包,我这里用的是mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64.tar.gz
下载完成后,使用ftp传送到服务器上面,然后分别在三台机器上解压
1
|
[root@sqltest01 u01] # tar -zxvf mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64.tar.gz
|
1
2
3
4
5
6
7
8
|
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/have_plugin_auth.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/kill_query.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/unsafe_binlog.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/have_multi_ndb.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/ipv6_clients.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/setup_fake_relay_log.inc
mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64/mysql-test/include/wait_for_slave_sql_error_and_skip.inc
......................................................................................................
|
解压后,里面包括了数据库文件以及集群软件
3、配置管理节点(sqltest03)
--将刚才解压的软件拷贝到指定位置/usr/local/mysql
1
2
3
|
[root@sqltest03 mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64] # pwd
/u01/mysql-cluster-gpl-7 .3.7-linux-glibc2.5-x86_64
[root@sqltest03 mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64] # mv * /usr/local/mysql/
|
--创建集群目录
1
2
3
4
5
6
7
8
9
|
[root@sqltest03 u01] # cd /usr/local/mysql
[root@sqltest03 mysql] # mkdir mysql-cluster
[root@sqltest03 mysql] # pwd
/usr/local/mysql
[root@sqltest03 mysql] # cp bin/ndb_mgm* /usr/local/bin/
[root@sqltest03 mysql] # cd /var/lib
[root@sqltest03 mysql] # mkdir mysql-cluster
[root@sqltest03 mysql] # cd mysql-cluster
[root@sqltest03 mysql] # vi config.ini
|
配置内容如下:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
[root@sqltest03 mysql-cluster]# cat config.ini
[ndbd default]
NoOfReplicas=1
DataMemory=2048M
IndexMemory=512M
[tcp default]
[ndb_mgmd]
hostname=192.168.9.243
datadir=/var/lib/mysql-cluster
NodeId=1
[ndbd]
hostname=192.168.9.241
datadir=/u01/mysql/data
NodeId=2
[ndbd]
hostname=192.168.9.242
datadir=/u01/mysql/data
NodeId=3
[mysqld]
hostname=192.168.9.241
NodeId=4
[mysqld]
hostname=192.168.9.242
NodeId=5
|
配置说明:
[ndbd default]
这部分是公共部分,对于每一个数据节点都有效,只需要配置一份
NoOfReplicas=1
数据镜像几份(各数据节点之间相互备份)
[tcp default]
针对每个数据节点及管理节点之间使用哪个端口进行通讯,在旧版本的NDB集群软件配置时,这个地方通常配置portnumber=2202但新版的NDB软件这里不需要配置,并且MySQL官方也强烈建议不要配置
[ndb_mgmd]
管理节点的配置部分(通常只有一个)。注意NodeId=1指明管理节点的节点ID为1,如果不指定,在启动集群时,会报错
hostname=192.168.9.243
指明管理节点的IP地址
datadir=/var/lib/mysql-cluster
指明集群管理日志存放的位置
[ndbd]
数据节点配置部分,有几个数据节点就配置几个[ndbd]
hostname=192.168.1.111
指明数据节点的IP地址
datadir=/u01/app/mysql/data
指明数据节点上的数据库文件存放的位置
NodeId=2
指明该数据节点在整个集群中的nodeid号(很重要)
[mysqld]
SQL节点配置部分,有几个SQL节点,就配置几个[mysqld]
到这里,就可以启动集群了
1
2
3
4
|
[root@sqltest03 bin] # pwd
/usr/local/bin
[root@sqltest03 bin] # ./ndb_mgmd -f /var/lib/mysql-cluster/config.ini
MySQL Cluster Management Server mysql-5.6.21 ndb-7.3.7
|
进入执行查看
1
|
[root@sqltest03 bin] # ndb_mgm
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
-- NDB Cluster -- Management Client --
ndb_mgm> show
Connected to Management Server at: localhost:1186
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 (not connected, accepting connect from 192.168.9.241)
id=3 (not connected, accepting connect from 192.168.9.242)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.9.243 (mysql-5.6.21 ndb-7.3.7)
[mysqld(API)] 2 node(s)
id=4 (not connected, accepting connect from 192.168.9.241)
id=5 (not connected, accepting connect from 192.168.9.242)
|
可以看到有两个节点,节点没有连接上
4、配置mysqld节点及存储节点(sqltest01,sqltest02)
--建立相应目录
1
2
3
|
[root@sqltest01 mysql] # mkdir -p /usr/local/mysql --用于存放刚才解压的文件,如mysql的bin目录等
[root@sqltest01 mysql] # mkdir -p /u01/mysql/data --用于存储数据文件(innodb)
[root@sqltest01 mysql] # chown -R mysql.mysql /u01
|
--将先前解压的文件拷贝
1
2
3
4
|
[root@sqltest01 mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64] # pwd
/u01/mysql-cluster-gpl-7 .3.7-linux-glibc2.5-x86_64
[root@sqltest01 mysql-cluster-gpl-7.3.7-linux-glibc2.5-x86_64] # mv * /usr/local/mysql/
[root@sqltest01 mysql] # chown -R mysql.mysql /usr/local/mysql/
|
--拷贝mysql.server
1
2
3
4
5
6
7
8
9
10
11
|
[root@sqltest01 support-files] # pwd
/usr/local/mysql/support-files
[root@sqltest01 support-files] # ls -lrt
total 32
-rw-r--r--. 1 mysql mysql 773 Oct 9 21:46 magic
-rwxr-xr-x. 1 mysql mysql 10880 Oct 9 22:42 mysql.server
-rwxr-xr-x. 1 mysql mysql 894 Oct 9 22:42 mysql-log-rotate
-rwxr-xr-x. 1 mysql mysql 1061 Oct 9 22:42 mysqld_multi.server
-rw-r--r--. 1 mysql mysql 1126 Oct 9 22:42 my-default.cnf
-rwxr-xr-x. 1 mysql mysql 1153 Oct 9 22:42 binary-configure
[root@sqltest01 support-files] # cp mysql.server /etc/rc.d/init.d/mysqld
|
--编辑环境变量
1
|
[root@sqltest01 tmp] # vi /etc/profile
|
添加如下:
1
|
PATH=/usr/local/mysql/bin:/usr/local/mysql/lib:$PATH
|
1
|
export PATH
|
1
|
[root@sqltest01 tmp] # source /etc/profile
|
--使修改生效
--配置my.cnf
1
|
[root@sqltest01 support-files] # cp my-default.cnf /etc/my.cnf
|
并对my.cnf进行配置,具体配置如下
1
2
3
4
5
6
7
|
[mysqld]
ndbcluster
basedir=/usr/local/mysql
datadir=/u01/mysql/data
port=3306
[mysql_cluster]
ndb-connectstring=192.168.9.243
|
--初始化节点数据库
1
|
[root@sqltest01 mysql] # scripts/mysql_install_db --basedir=/usr/local/mysql --datadir=/u01/mysql/data
|
执行完这条命令以后,数据库的数据文件(包括mysql,test , performance_schema等数据库),就安装到相应的位置了,可以使用了
在两个节点都执行上面的步骤即可
5、在两个点启动
1
|
[root@sqltest01 mysql] # ndbd --initial
|
1
2
|
2014-12-24 17:55:57 [ndbd] INFO -- Angel connected to '192.168.9.243:1186'
2014-12-24 17:55:57 [ndbd] INFO -- Angel allocated nodeid: 2
|
第一次启动时,需要加--initial来初始化数据节点,第二次启动时,就不需要这个参数了
在管理节点查看,可以看到第一个节点已经连接
1
|
ndb_mgm> show
|
1
2
3
4
5
6
7
8
9
10
|
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 @192.168.9.241 (mysql-5.6.21 ndb-7.3.7, starting, Nodegroup: 0) --表明已经连接上了
id=3 (not connected, accepting connect from 192.168.9.242)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.9.243 (mysql-5.6.21 ndb-7.3.7)
[mysqld(API)] 2 node(s)
id=4 (not connected, accepting connect from 192.168.9.241)
id=5 (not connected, accepting connect from 192.168.9.242)
|
启动mysqld
1
2
|
[root@sqltest01 mysql] # cd /usr/local/mysql/bin
[root@sqltest01 bin] # ./mysqld_safe --user=mysql
|
1
2
|
141224 17:59:50 mysqld_safe Logging to '/u01/mysql/data/sqltest01.err'.
141224 17:59:51 mysqld_safe Starting mysqld daemon with databases from /u01/mysql/data
|
启动数据库时,第一次初始化使用的root,而这次使用mysql,需要对/u01/mysql/data权限进行配置,否则报不可读写
再次在管理节点查看
1
|
ndb_mgm> show
|
1
2
3
4
5
6
7
8
9
|
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 @192.168.9.241 (mysql-5.6.21 ndb-7.3.7, Nodegroup: 0, *)
id=3 (not connected, accepting connect from 192.168.9.242)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.9.243 (mysql-5.6.21 ndb-7.3.7)
[mysqld(API)] 2 node(s)
id=4 @192.168.9.241 (mysql-5.6.21 ndb-7.3.7)
|
--表明已经连接上了
最后把第二节点也启动,再次从管理节点检查
1
|
ndb_mgm> show
|
1
2
3
4
5
6
7
8
9
10
|
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 @192.168.9.241 (mysql-5.6.21 ndb-7.3.7, Nodegroup: 0, *)
id=3 @192.168.9.242 (mysql-5.6.21 ndb-7.3.7, Nodegroup: 1)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.9.243 (mysql-5.6.21 ndb-7.3.7)
[mysqld(API)] 2 node(s)
id=4 @192.168.9.241 (mysql-5.6.21 ndb-7.3.7)
id=5 @192.168.9.242 (mysql-5.6.21 ndb-7.3.7)
|
6、在两个节点测试
1
|
[root@sqltest01 ~] # mysql -uroot -p
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
Enter password:
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 2
Server version: 5.6.21-ndb-7.3.7-cluster-gpl MySQL Cluster Community Server (GPL)
Copyright (c) 2000, 2014, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql> create database mydb1;
Query OK, 1 row affected (0.07 sec)
mysql> use mydb1;
Database changed
mysql> create table mytb1(id int,birthdate datetime,pername char(10)) engine=ndbcluster;
Query OK, 0 rows affected (0.19 sec)
|
1
|
mysql> insert into mytb1( id ,birthdate,pername) values(1, '2013-01-23 09:45:10' , 'pengzitong' );
|
1
2
3
4
5
|
Query OK, 1 row affected (0.00 sec)
mysql> commit;
Query OK, 0 rows affected (0.00 sec)
mysql> insert into mytb1(id,birthdate,pername) values(2,'2007-07-09 09:45:10','pengzixin');
Query OK, 1 row affected (0.00 sec)
|
1
|
mysql> commit;
|
1
|
Query OK, 0 rows affected (0.00 sec)
|
在第二节点检查
1
|
[root@sqltest02 ~] # mysql -uroot -p
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
Enter password:
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 2
Server version: 5.6.21-ndb-7.3.7-cluster-gpl MySQL Cluster Community Server (GPL)
Copyright (c) 2000, 2014, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
mysql> use mydb1
mysql> select * from mytb1;
+------+---------------------+------------+
| id | birthdate | pername |
+------+---------------------+------------+
| 1 | 2013-01-23 09:45:10 | pengzitong |
| 2 | 2007-07-09 09:45:10 | pengzixin |
+------+---------------------+------------+
|
7、集群停止
要想关闭 Cluster,可在MGM节点所在的机器上,在Shell中简单地输入下述命令:
1
|
[root@sqltest03 bin] # /usr/local/mysql/ndb_mgm -e shutdown
|
运行以下命令关闭SQL节点的mysqld服务:
1
|
[root@sqltest01 bin] # /usr/local/mysql/bin/mysqladmin -uroot shutdown
|