MySQL high availability practical solution MHA

One 、 About MHA I have to understand

Use MHA It can liberate our hands to a greater extent , Accomplish more with fewer instructions ,MHA Mainly able to do the following things ：

• Automatic in MASTER Elect a new one after the outage SLAVE As MASTER, Ensure that the service is not interrupted .

• Automatic in MASTER After the outage, all those who have not been elected as new MASTER Of SLAVE Point back to the new MASTER And start replication .

• Automatic in MASTER Send an alarm email to the database administrator after the outage Automatically VIP Drift service , Ensure that the service operation will not be suspended .

• MHA The minimum construction conditions are 1 Lord 2 from , And it must be an independent server , You cannot build multiple instances on a single machine .

Two 、 Architecture demonstration

MHA It's actually a collection of software , Its software is divided into 2 part ：

• Manager Software
• Node Software

Node The software must be installed in all MHA Node , and Manager The software only needs to be installed on the management node .

Different software consists of different toolkits , As shown below ：

-- Master
masterha_manager            -  Used to start MHA
masterha_check_ssh          -  Used for inspection MHA Of SSH Configuration situation 
masterha_check_repl         -  Used for inspection MHA Master-slave replication of 
masterha_master_monitor     -  Used for inspection Master Whether the node is down 
masterha_check_status       -  Used to check the current MHA Operating state 
masterha_master_switch      -  For automatic fault recovery 
masterha_conf_host          -  Used to add or delete Manager Configured in ser