Shulou(Shulou.com)06/01 Report--

First, based on the disadvantages of ordinary master-slave replication, MHA can automatically complete the database failover operation within 10 to 30 seconds, and in the process of failover, MHA can ensure the consistency of data to the greatest extent to achieve high availability in the real sense.

II. Structure

III. Software introduction

MHA software consists of two parts, the Manager toolkit and the Node toolkit.

Manager toolkit

Masterha_check_ssh # check the ssh- key ^ of MHA

Masterha_check_repl # check master-slave replication

Masterha_manger # launch MHA

Masterha_check_status # detects the running status of MHA ^

Masterha_mast er_monitor # check whether master is down

Masterha_mast er_switch # Manual failover-

Masterha_conf_host # add server double interest one manually

Masterha_secondary_check # establish a TCP connection from a remote server v

Masterha_stop # stop MHA

The Node toolkit mainly includes the following tools:

Save_binary_1ogs # saves the binlog of the down master

Apply_diff_relay_logs # identifies differences in relay log

Filter_mysqlbinlog # prevent rollback event-MHA no longer uses this tool

Purge_relay_logs # clearing relay log one does not block SQL threads

Fourth, the advantages of MHA

1. Automatic failover

2. There is no data inconsistency when the main database crashes.

3. There is no need to make major changes to the current mysql environment

4. There is no need to add additional servers

5. Excellent performance, able to work with semi-synchronous and asynchronous replication frameworks

6. As long as the storage engine mha supported by replication supports it.

5. Environmental description

Db01 host (master)

[root@db01 ~] # cat / etc/redhat-release

CentOS release 6.9 (Final)

[root@db01 ~] # uname-r

2.6.32-696.el6.x86_64

[root@db01 ~] # / etc/init.d/iptables status

Iptables: Firewall is not running.

[root@db01 ~] # getenforce

Disabled

[root@db01] # hostname-I

10.0.0.51 172.16.1.51

Db02 host (slave1)

[root@db02 ~] # cat / etc/redhat-release

CentOS release 6.9 (Final)

[root@db02 ~] # uname-r

2.6.32-696.el6.x86_64

[root@db02 ~] # / etc/init.d/iptables status

Iptables: Firewall is not running.

[root@db02 ~] # getenforce

Disabled

[root@db02] # hostname-I

10.0.0.52 172.16.1.52

Db03 hosts (slave1,MHA Manages, Atlas nodes)

[root@db03 ~] # cat / etc/redhat-release

CentOS release 6.9 (Final)

[root@db03 ~] # uname-r

2.6.32-696.el6.x86_64

[root@db03 ~] # / etc/init.d/iptables status

Iptables: Firewall is not running.

[root@db03 ~] # getenforce

Disabled

[root@db03] # hostname-I

10.0.0.53 172.16.1.53

Mysql 5.6.38 is newly installed on all three servers:

[root@db01] # mysql-- version

Mysql Ver 14.14 Distrib 5.6.38, for Linux (x86 / 64) using EditLine wrapper

Master-slave replication configuration based on GTID

Binlog should be enabled for both the master library and the slave library

The master library and slave library server-id must be different

There must be a master-slave replication user.

5.1. Configure master-slave replication

Db01 my.cnf file

[mysqld]

Basedir=/application/mysql

Datadir=/application/mysql/data

Socket=/tmp/mysql.sock

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

Log-bin=/application/mysql/mysql-bin

Binlog_format=row

Secure-file-priv=/tmp

Server-id=51

Skip-name-resolve # skip domain name resolution

Gtid-mode=on # enables the gtid type, otherwise it is a normal replication schema

Enforce-gtid-consistency=true # enforces consistency in GTID

Whether log-slave-updates=1 # slave updates are logged (5.6 required)

Relay_log_purge = 0

[mysql]

Socket=/tmp/mysql.sock

Db02 my.cnf file

[root@db02 ~] # cat / etc/my.cnf

[mysqld]

Basedir=/application/mysql

Datadir=/application/mysql/data

Socket=/tmp/mysql.sock

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

Log-bin=/application/mysql/mysql-bin

Binlog_format=row

Secure-file-priv=/tmp

Server-id=52

Skip-name-resolve

Gtid-mode=on

Enforce-gtid-consistency=true

Log-slave-updates=1

Relay_log_purge = 0

[mysql]

Socket=/tmp/mysql.sock

Db03 my.cnf file

[root@db03 ~] # cat / etc/my.cnf

[mysqld]

Basedir=/application/mysql

Datadir=/application/mysql/data

Socket=/tmp/mysql.sock

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

Log-bin=/application/mysql/mysql-bin

Binlog_format=row

Secure-file-priv=/tmp

Server-id=53

Skip-name-resolve

Gtid-mode=on

Enforce-gtid-consistency=true

Log-slave-updates=1

Relay_log_purge = 0

Skip-name-resolve

[mysql]

Socket=/tmp/mysql.sock

Create replication users (51 as master nodes, 52 and 53 as slaves)

GRANT REPLICATION SLAVE ON *. * TO repl@'172.16.1.%' IDENTIFIED BY '123'

Enable replication from the library

Change master to

Master_host='172.16.1.51'

Master_user='repl'

Master_password='123'

MASTER_AUTO_POSITION=1

Start replication from the library

Start slave

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.