MySQL Cluster集群负载配置
一、软件包
a) MySQL-ndb-management-5.0.90-1.i386.rpm
b) MySQL-ndb-management-5.0.90-1.i386.rpm
c) mysql-5.0.27.tar.gz
二、安装［管理节点］
a) rpm -ivh MySQL-ndb-tools-5.0.90-1.i386.rpm
b) rpm -ivh MySQL-ndb-management-5.0.90-1.i386.rpm
c) mkdir /var/lib/mysql-cluster
d) cd /var/lib/mysql-cluster
e) touch config.ini
[NDBD DEFAULT]
NoOfReplicas=2
[MYSQLD DEFAULT]
[NDB_MGMD DEFAULT]
[TCP DEFAULT]
# Managment Server
[NDB_MGMD]
HostName=192.168.1.67
# Storage Engines
[NDBD]
HostName=192.168.1.91
DataDir=/usr/local/mysql/data
#DataDir=/var/lib/mysql-cluster
[NDBD]
HostName=192.168.1.92
DataDir=/usr/local/mysql/data
#DataDir=/var/lib/mysql-cluster
[MYSQLD]
[MYSQLD]
f) /usr/bin/ndb_mgm �Cinitial
g) /usr/bin/ndb_mgmd shutdown
h) /usr/bin/ndb_mgmd -f /var/lib/mysql-cluster/config.ini［启动管理节点］
i) /usr/sbin/ndb_mgmd -f /etc/ndb_mgmd.cnf
三、安装［数据节点］
a) Cd /var/local/software
b) Tar zxvf mysql-5.0.27.tar.gz
c) Cd mysql-5.0.27
d) ./configure --prefix=/usr/local/mysql --localstatedir=/usr/local/mysql/data --disable-maintainer-mode --with-mysqld-user=mysql --with-unix-socket-path=/tmp/mysql.sock --without-comment --without-debug --with-bench --with-ndbcluster;
e) Make & make install
f) Finger mysql
g) Useradd mysql
h) cd scripts
i) ./mysql_install_db
j) chown -R mysql:mysql /usr/local/mysql
k) chown -R mysql:mysql /usr/local/mysql/data
l) cd /usr/local/src/mysql-5.0.27/support-files/
m) cp my-medium.cnf /etc/my.cnf
n) killall mysqld
o) /usr/local/mysql/bin/mysqld_safe --user=mysql --skip-grant-tables --skip-networking &
# /usr/local/mysql/bin/mysql -u root mysql
mysql> UPDATE user SET Password=PASSWORD('123456') where USER='root';
mysql> FLUSH PRIVILEGES;
mysql> quit
修改root用户远程访问的权限：
# killall mysqld
# /usr/local/mysql/bin/mysqld_safe �Cuser=mysql &
# /usr/local/mysql/bin/mysql -u root �Cp123456 mysql
mysql> GRANT ALL PRIVILEGES ON *.* TO 'root'@'%' IDENTIFIED BY 'root' WITH GRANT OPTION;
mysql> FLUSH PRIVILEGES;
mysql> quit
vi /etc/my.cnf
[mysqld]
ndbcluster
ndb-connectstring=192.168.1.67
[mysql_cluster]
ndb-connectstring=192.168.1.67
killall mysqld
/usr/local/mysql/bin/mysqld_safe &
/usr/local/mysql/libexec/ndbd �Cinitial（初始化数据节点）
四、 查看管理节点
a) /usr/bin/ndb_mgmd shutdown
b) /usr/bin/ndb_mgmd -f /var/lib/mysql-cluster/config.ini
c) /usr/bin/ndb_mgm
-- NDB Cluster -- Management Client --
ndb_mgm> show
Connected to Management Server at: localhost:1186
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 @192.168.1.91 (Version: 5.0.27, Nodegroup: 0, Master)
id=3 @192.168.1.92 (Version: 5.0.27, Nodegroup: 0)
[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.1.67 (Version: 5.0.90)
[mysqld(API)] 2 node(s)
id=4 @192.168.1.92 (Version: 5.0.27)
id=5 @192.168.1.91 (Version: 5.0.27)
ndb_mgm>
以上为正常显示
五、测试节点同步
a) 在其中一数据节点中/usr/local/mysql/bin/mysql �Curoot �Cp
b) create database member;
use member ;
create table admin (id int ) engine=ndbcluster ;
insert into admin values(1) ;
c) 在另一数据节点中/usr/local/mysql/bin/mysql �Curoot �Cp
d) Create database member;
e) Use member;
f) flush tables ;
g) Show tables;
h) 可以看到admin表已同步
i) Select * from admin
j) 可以看到有一条记录已同步
k) 集群正常
六、相关命令
a) /usr/local/mysql/libexec/ndbd �Cinitial（初始化数据节点）
b) /usr/bin/ndb_mgmd shutdown(关闭管理节点)
c) /usr/bin/ndb_mgmd -f /var/lib/mysql-cluster/config.ini(启动管理节点)
d) /usr/bin/ndb_mgm查看集群情况
e) Ps aux|grep mysql查看mysql是否启动
f) Ps aux|grep ndb查看管理节点是否启动
g) Ndb_mgm �Ce “show”;
h) Ndb_mgm>all stauts;
i) Ndb_mgm >2 status;(2为节点编号)
j) Ndbd �Cno-wait-nodes=2,3
k) Mysql>Show engine ndb status;
l) Ndb_mgm>clusterlog info;
m) Mysql>show status like ‘ndb%’;
n) Mysql>show warnings;
o) Ndb_mgm>2 stop;
p) SHOW ENGINES;
Vi /etc/ndb_mgmd.cnf
[NDBD DEFAULT]
NoOfReplicas=2
MaxNoOfConcurrentOperations= 10000
DataMemory=1800M
IndexMemory=200M
MaxNoOfAttributes=500000[无法创建表问题]
MaxNoOfTables=1760
MaxNoOfUniqueHashIndexes=5000
MaxNoOfOrderedIndexes=5000
MaxNoOfTriggers=4000
TimeBetweenLocalCheckpoints=2
NoOfFragmentLogFiles=100
DataDir= /var/lib/mysql-cluster
[MYSQLD DEFAULT]
[NDB_MGMD DEFAULT]
DataDir= /var/lib/mysql-cluster
[TCP DEFAULT]
# Managment Server
[NDB_MGMD]
HostName=192.168.1.67
# Storage Engines
[NDBD]
HostName=192.168.1.91
MaxNoOfAttributes=500000
DataDir=/usr/local/mysql/data
#DataDir=/var/lib/mysql-cluster
[NDBD]
HostName=192.168.1.92
MaxNoOfAttributes=500000
DataDir=/usr/local/mysql/data
#DataDir=/var/lib/mysql-cluster
[MYSQLD]
[MYSQLD]
七、IPVSADM安装
a) Rpm �Civh ipvsadm-1.24-8.1.i386.rpm
b) Vi /opt/ipvsadm.sh
#!/bin/bash
ipvsadm -C
ipvsadm -A -t 59.64.28.94:3306 -s rr
ipvsadm -a -t 59.64.28.94:3306 -r 59.64.28.91:3306 -g -w 1
ipvsadm -a -t 59.64.28.94:3306 -r 59.64.28.92:3306 -g -w 1
/etc/rc.d/init.d/ipvsadm save
c) Chmod 755 ipvsadm.sh
d) Service ipvsadm restart
e) Vi /etc/sysconifg/ipvsadm
八、Keepalived安装(监测心跳)
a) Tar zxvf keepalived-1.1.15.tar.gz
b) ln -s /usr/src/kernels/2.6.18-164.el5-PAE-i686 /usr/src/linux
c) Cd keepalived-1.1.15
d) ./configure --prefix=/usr/local/keepalive/ --with-kernel-dir=/usr/src/linux
e) Make
f) Make install
g) cd /usr/local/keepalive/
h) cp /usr/local/keepalive/etc/rc.d/init.d/keepalived /etc/rc.d/init.d/
i) cp /usr/local/keepalive/etc/sysconfig/keepalived /etc/sysconfig/
j) mkdir /etc/keepalived
k) cp /usr/local/keepalive/etc/keepalived/keepalived.conf /etc/keepalived/
l) cp /usr/local/keepalive/sbin/keepalived /usr/sbin/
m) chkconfig --add keepalived
n) chkconfig keepalived on
o) service keepalived restart
p) vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
}
notification_email_from [email protected]
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id LVS_DEVEL
}
vrrp_instance VI_1 {
state MASTER
interface eth1
lvs_sync_daemon_interface eth1
virtual_router_id 51
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
59.64.28.94 dev eth1 label eth1:1(虚拟IP)
}
}
virtual_server 59.64.28.94 3306 {
delay_loop 6
lb_algo rr
lb_kind DR
nat_mask 255.255.255.0
persistence_timeout 60
protocol TCP
real_server 59.64.28.91 3306 {
weight 1
TCP_CHECK {
connect_timeout 10
nb_get_retry 3
delay_before_retry 3
connect_port 3306
}
}
real_server 59.64.28.92 3306 {
weight 1
TCP_CHECK {
connect_timeout 10
nb_get_retry 3
delay_before_retry 3
connect_port 3306
}
}
}
q) service keepalived restart
r) ipvsadm
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 59.64.28.94:mysql rr persistent 60
-> 59.64.28.92:mysql Route 1 0 0
-> 59.64.28.91:mysql Route 1 0 0
s) 应用程序数据库连接主机改成59..64.28.94测试。
九、相关命令
a) service ipvsadm restart
b) service keepalived restart
c) ipvsadm
d) vi /usr/local/keepalive/etc/keepalived/keepalived.conf[/etc/keepalived/keepalived.conf]