Step-by-Step Guide to Deploying MHA for MySQL High Availability

Prerequisite

MHA does not automatically create master‑slave topology, so you must set up a MySQL master‑slave environment manually. Ensure all slaves are not read‑only and have binary logging enabled; the designated failover slave must have binlog enabled and be read‑write.

Architecture

System Environment

# cat /etc/redhat-release

CentOSrelease 6.6 (Final) # uname -rm 2.6.32-504.el6.x86_64

Create MHA User

Run on the master server:

mysql> grant all privileges on *.* to 'mha'@'192.168.56.%' identified by '123456';

Create Symlinks (if MySQL not installed via yum)

# ln -s /application/mysql/bin/mysqlbinlog /usr/bin/mysqlbinlog

# ln -s /application/mysql/bin/mysql /usr/bin/mysql

Configure SSH Public‑Key Authentication

Generate a DSA key pair on the manager host: # ssh-keygen -t dsa Distribute the public key to all MySQL hosts.

Configure Hosts File

Deploy MHA Node

Install MHA Node on every MySQL server (both master and slaves). Example on the manager host: # mkdir /softs Then install the node package (see image for detailed steps).

Deploy MHA Manager

The manager runs on a designated server (can be a slave). Install the manager package:

Standardize the MHA directory layout:

Configure app1.cnf (example screenshots):

Set up the global configuration file:

Check Configuration

Run these checks on the manager host.

SSH Connectivity

Replication Status

Start Manager

Check Manager Status

Master Failover Test

Simulate master failure and observe automatic promotion:

Verify the new master (e.g., mysql-slave02) shows the expected data and replication state:

After verification, the MHA cluster provides seamless MySQL high availability.