MHA Mysql achieve a master multi-slave architecture

MHA profile:

MHA (Master High Availability) currently in terms of high availability MySQL is a relatively mature solution that youshimaton by the Japanese company DeNA (Facebook now working in the company) to develop, is a good case as a failure MySQL high availability environment and the main switch from an elevated high availability software. In MySQL failover process, the MHA can be done automatically in 30 seconds from 0 to the fault switching operation of the database, and performing the handover procedure failure, the MHA to ensure consistency of the data to the maximum extent, in order to achieve real availability of meaning.
MHA also offers online master library function switch can be switched safely to the main library currently running a new main library (library by lifting from the main library), it can be completed within about 0.5-2 seconds.

MHA works:

Simply draw a bit

Preparing the Environment

centos7 version of the above

cat /etc/redhat-release

Kernel above 3.10

uname -r

Machine name	IP configuration	Server Role	Remark
Manager	10.0.0.41	Manager	For the monitoring and management
Master	10.0.0.42	Primary database server	Open bin-log relay-log close relay_log_purge
Slave1	10.0.0.43	From the database server	Open bin-log relay-log close relay_log_purge
Slave2	10.0.0.44	From the database server	Open bin-log relay-log close relay_log_purge

In order to set free secret login convenience, within etc / hosts file on each node in the configuration of add about content:

10.0.0.41 node1.keer.com node1
10.0.0.42 node1.keer.com node2
10.0.0.43 node1.keer.com node3
10.0.0.44 node1.keer.com node4

As FIG.

From a multi-master configuration

A multi-master from what I can learn, you can do your own configuration from a multi-master and very detailed

1. Configuring ssh-free dense Login

ssh-keygen -t rsa

After executing the commands you need to enter a carriage return himself three times

ll -a

cd .ssh

ssh-copy-id [email protected]
ssh-copy-id [email protected]
ssh-copy-id [email protected]

1. On Manager server
   
2. On the Master Server
   
3. Slave1 on the server

4. On Slave2 server

The above operation begins configure ssh login-free dense, each node needs to perform again, in addition to IP needs a change.

5. MHA installation package

Manager requires two packages below
Three other nodes only need to install a packet as in FIG.

Required packages can go to my Web site to download: the installation package of MHA
MHA and three other nodes require installation package

6. Set MHA Management ProfilesMHANode operation

mkdir /etc/mha_master
vim /etc/mha_master/mha.cnf

Configuration files are as follows

[server default]
user=mhaadmi  #Master设置的管理用户
password=123456 #Master设置的管理用户的密码
manager_workdir=/etc/mha_master/app1
manager_log=/etc/mha_master/manager.log
remote_workdir=/mydata/mha_master/app1
ssh_user=root  #ssh的免密认证
repl_user=root  #数据库授权的用户
repl_password=123456  #数据库授权的的密码
ping_interval=1
[server1]
hostname=10.0.0.42 #Master的ip
ssh_port=22
candidate_master=1
[server2]
hostname=10.0.0.43  #Slave1的ip
ssh_port=22
candidate_master=1
[server3]
hostname=10.0.0.44  #Slave2的ip
ssh_port=22
candidate_master=1

7. Ssh trust communication between nodes detected configuration is okOperation on MHA

masterha_check_ssh -conf=/etc/mha_master/mha.cnf

8. Connection configuration parameters MySQL replication cluster management to check whether the OK to execute on the Master node

grant all on *.* to 'mhaadmi'@'10.0.%.%' identified by '123456';
flush privileges;

masterha_check_repl -conf=/etc/mha_master/mha.cnf

Below the last row of executing the display MySQL Replication Health is OK. To

9. Start MHA
   executed on MHA node

 nohup masterha_manager -conf=/etc/mha_master/mha.cnf &> /etc/mha_master/manager.log &

FIG been executed as follows

10. Look at the state after the start

masterha_check_status -conf=/etc/mha_master/mha.cnf

FIG been executed as follows

11. MHA test failover
    in the Master node closed Mariadb service, analog Master node goes down

systemctl stop mariadb

Check node status in Manager

masterha_check_status -conf=/etc/mha_master/mha.cnf

FIG been executed as follows

12. View Slave2 master-slave synchronization

Now Slave2 show the Lord's IP is an IP Slave1 then show success