Mysql+drbd+heartbeat implementation of MySQL high-availability dual standby (mysql+heartbeat)

Deploying mysql*******************************************

Yum-y install gcc gcc-c++ gcc-g77 autoconf automake zlib* fiex* libxml* ncurses-devel libmcrypt* libtool-ltdl-devel* Mak E CMake bison*

Useradd mysql-d/usr/local/mysql-s/sbin/nologin

TAR-ZXVF mysql-5.1.72.tar.gz

CD mysql-5.1.72.tar.gz

./configure--prefix=/usr/local/mysql/--with-charset=utf8

Make && make install

Chown-r Mysql.mysql/usr/local/mysql

Cd/usr/local/mysql/scripts

./mysql_install_db--user=mysql--basedir=/usr/local/mysql/--datadir=/data/mysql

Copy configuration file to/etc directory

[email protected] mysql-5.1.72]# CP support-files/my-medium.cnf/etc/my.cnf

Copy startup script to resource Directory

CP Support-files/mysql.server/etc/rc.d/init.d/mysqld

Add MySQL Service

[[email protected] mysql-5.1.72] #chmod +x/etc/rc.d/init.d/mysqld

Mysqladmin–u root password 123456//Set Roo password

10. Start MySQL

[[email protected] mysql-5.1.72] #service mysqld start

#配置库文件搜索路径

#echo "/usr/local/mysql/lib/mysql" >>/etc/ld.so.conf

#ldconfig

#添加/usr/local/mysql/bin to the environment variable path

#export path= $PATH:/usr/local/mysql/bin

Cp-r/data/mysql

Chown-r Mysql.mysql/data/mysql

# #编辑Mysql的配置文件/ETC/MY.CNF

[Email protected] ~]# sed-n "/^[^#]/p"/etc/my.cnf

[Mysqld]

Server-id=1 # #保证全局唯一

Log-bin=mysql-bin# #开启记录二进制日志的功能

Relay-log=mysql-relay-bin

datadir=/data/mysql/

Socket=/data/mysql/mysql.sock

User=mysql

Symbolic-links=0

[Mysqld_safe]

Log-error=/var/log/mysqld.log

Pid-file=/var/run/mysqld/mysqld.pid

To manage MySQL boot to heartbeat, create a soft link to the MySQL startup script in the Heartbeat resource script directory

Ln-s/etc/init.d/mysqld/etc/ha.d/resource.d/mysqld

# # # # #重启mysql服务 #######

Service mysqld Restart

############# #配置mysql的主从复制 ####################

Add an account for replication on master

Mysql>grant REPLICATION SLAVE on * * to ' repl ' @ ' 10.0.0.2 ' identified by ' repl123456 ';

To back up the data on master, first execute the following statement:

Mysql>flush TABLES with READ LOCK;

Mysql>reset master;

Do not exit the terminal, or the lock will fail, in the case of not exiting the terminal, and then open a terminal directly packaged compressed data files or use the Mysqldump tool to export data.

Cd/data/mysql

Tar zcvf mysql.tar.gz/data/mysql

SCP mysql.tar.gz [Email protected]:/root/

After the data transfer is complete, execute the following command on the master terminal that just executed the flushtables with READ LOCK command:

Mysql> UNLOCK TABLES;

To view the status of master:

Mysql>show Master Status\g #以下面列出的配置, set the Master_log_file value of the slave, and the Master_log_pos value.

Setting up the slave host

Turn on master and slave synchronization and execute the following statement on slave:

Mysql>change MASTER to master_host= ' 10.0.0.1 ',

->master_user= ' Repl ',

->master_password= ' repl123456 ',

->master_log_file= ' mysql-bin.000001 ',

->master_log_pos=106;

When you are done, turn on slave

Mysql>slave start;

View slave status

Mysql>show slave Status\g

When both slave_io_running and slave_sql_running are yes, the configuration is successful.

A few problems to be noted ********************************

1) If the Log-bin, relay-log parameters are defined in the MY.CNF, then ensure that the definitions are independent of the hostname,

Because if the log file names of these two classes are related to the hostname, the switchover process causes the slave host to not continue to synchronize, which can be set as follows:

Log-bin=mysql-bin

Relay-log=mysql-relay-bin

Ensure that the names of two files on both hosts are the same

2) To ensure that MySQL can pass start\stop\status three parameters, the default heartbeat is in the LSB style,

A return value that contains OK or running indicates that the resource is healthy, that the return value contains stopped or no indicates that the resource is unhealthy, and that other resource scripts are similar.

3) do not set MySQL to start automatically when the machine restarts, mysqld as a resource of heartbeat, Heartbeat will be unified management.

4) Ensure that each MySQL my.cnf configuration file is identical to other configuration items except Server-id.

Heartbeat Deployment ****************************************

The heartbeat card for both hosts is eth1

Node1 's heartbeat IP is: 10.0.0.1

Node2 's heartbeat IP is: 10.0.0.2

Install Heartbeat:

1. Create users and Groups first

Groupadd-g haclient

Useradd-u 2000-g haclient Hacluster

Packages required to install heartbeat:

Yum-y Install pkgconfig glib2-devel e2fsprogs libxml2* PyXML

Recompile and install Libnet

TAR-ZXVF libnet-1.1.2.1.tar.gz-c./

./configure

Make && make install

Installing Heartbeat

TAR-ZXVF heartbeat-2.1.2.tar.gz-c./

./configureme Configure

make&& make Install

Configure Heartbeat

There are three configuration files required: HA.CF, Haresources, Authkeys.

These three configuration files need to be under the/ETC/HA.D directory, but the default is no three files, so you want to:

Copy storage Path/heartbeat-2.1.2/doc/ha.cf/etc/ha.d/

Copy storage Path/heartbeat-2.1.2/doc/haresources/etc/ha.d/

Copy storage Path/heartbeat-2.1.2/doc/authkeys/etc/ha.d/

The permission to modify Authkeys is 600

chmod 600/etc/ha.d/authkeys

Configure HA.CF

[Email protected] ~]# sed-n "/^[^#]/p"/etc/ha.d/ha.cf

LogFile/var/log/ha-log #指定heartbeat日志的位置

Logfacilitylocal0

KeepAlive 1 #每秒检测心跳

Deadtime#备用节点15s内没有检测到主机心跳 to confirm the failure of each other

Warntime 5#警告5次

Initdead#守护进程启动30s后, start service resources

Ucast eth1 10.0.0.1#另一台主机的eth1和eth0的ip地址 with two different networks to detect heartbeat availability

Ucast eth0 10.2.16.254

Auto_failback off #当主节点切换到备节点之后, when the primary node returns to normal, no failback is done because the cost of switching mysql-master is high

Respawn Hacluster/usr/lib64/heartbeat/ipfail

Respawn hacluster/usr/lib64/heartbeat/dopd #开启dopd功能

Apiauth Ipfail gid=haclient Uid=hacluster

Apiauth DOPD gid=haclient Uid=hacluster

Node Node1 defines the host name of the nodes

Node Node2 defines the host name of the nodes

Configuring Resource Files Haresources

/etc/ha.d/haresources

Node1 drbddisk::r0 filesystem::/dev/drbd0::/data mysqld ipaddr::10.2.16.250/24/eth0

Node1 is the host name of the master node

Drbddisk is a script to manage DRBD, default heartbeat not, need to copy this script to/etc/ha.d/resource.d/drbddisk (at the end of this script, copy and paste directly)

R0 is the startup resource for DRBD, "drbddisk::r0" can switch the host for DRBD to primary node or secondary node, only nodes with a status of primary can mount the DRBD partition

The Drbddisk script is equivalent to performing a DRBDADM primary r0 or Drbdadm secondary r0 operation to change and mount the resource role of DRBD

Mysqld means the MySQL service that started the machine

Ipaddr::10.2.16.250/24/eth0 indicates IP drift for virtual IP 250

Configuring Authkeys Certification Files

/etc/ha.d/authkeys

Auth 1

1 SHA1 ha_cluster #使用sha1验证, Password: ha_cluster

#将drbddisk脚本加执行权限:

Chmod+x/etc/ha.d/resource.d/drbddisk

After setting the master node, copy the HA.CF haresources Authkeys and Drbddisk file to the standby node directly.

SCP ha.cf haresources authkeys resources.d/drbddisk [email protected]:/root/

After moving to the HA.D directory on the standby node, modify the Ucast value of the HA.CF to be the primary node IP, the other does not change

Verify that the MySQL service is turned off, and that the heartbeat service is turned on on both hosts

Service Heartbeat Start

Observe the startup process of heartbeat by looking at/var/log/ha-log

This article is from the "Fate" blog, make sure to keep this source http://czybl.blog.51cto.com/4283444/1426063