MHA Building--percona Server

aVersion Information

System Environment: CentOS 6.5 X64

MHA software version:

Mha4mysql-manager-0.56.tar.gz

Mha4mysql-node-0.56.tar.gz

MySQL Version:

Percona-server-5.6.22-rel71.0-726.linux.x86_64.tar.gz

twoInstallation Steps

Prerequisites,master-slave architecture (2.8.19 supports a master-one from the architecture, this article does not involve the construction of master-slave architecture)

(i) InstallationNodenode

(This configuration is in each operation on the DB node)

1.Installing the Software support Package

Yum-y Install Perl-extutils-makemakerperl-extutils-cbuilder Perl-cpan perl-dbi perl-dbd-mysql

2.Installing the Software

Perl makefile.pl

Make && make install

(ii) installationManager

The manager will also need to install node if it is deployed independently on a separate server , and install node after installing the Manager

1.Installing the Software support Package

Yum-y Install Perl-time-hiresperl-config-tiny

Cpan-i Parallel::forkmanager # middle requires interactive operation input Yes( need to resolve how to do non-interactive installation for automatic installation)

Cpan-i Log::D ispatch

2.Editing a configuration file

# MKDIR/ETC/MASTERHA

# mkdir-p/masterha/app1

# CP samples/conf/*/etc/masterha/

# CAT/ETC/MASTERHA/APP1.CNF

[Server default]

Manager_workdir=/masterha/app1

Manager_log=/masterha/app1/manager.log

User=mha

password=123456

Ssh_user=root

Repl_user=repl

repl_password=654321

Ping_interval=1

[Server1]

hostname=172.16.52.130

Candidate_master=1 # Set the elevation level, if each is 1, the failover main library is promoted in the configuration order

Master_binlog_dir= "/usr/local/mysql/data" # binlog Storage path for the Master library

[Server2]

hostname=172.16.52.131

Candidate_master=1

Master_binlog_dir= "/usr/local/mysql/data"

[Server3]

hostname=172.16.52.132

Candidate_master=1

Master_binlog_dir= "/usr/local/mysql/data"

Description

1)master_binlog_dir binlog log storage path corresponding to the main library

(c) no secret landing

Configuration Manager and node ssh each other without a password login, each DB needs to be configured

Ssh-keygen

Ssh-copy-id-i IP

(iv) database authorization

for each DB for authorization,( each node includes the manager needs to be configured )

Grant all on * * [email protected] ' 172.16.52.% ' identified by ' 123456 ';

Flush privileges;

in order to ensure subsequent rapid switching, the same user Authorization for replication on DB:

GRANT REPLICATION SLAVE on * * to ' repl ' @ ' 172.16.52.% ' identified by ' 654321 ';

Flush privileges;

(v) TestingSshConnection

Masterha_check_ssh--conf=/etc/masterha/app1.cnf

(vi) test master-Slave synchronization

Masterha_check_repl--conf=/etc/masterha/app1.cnf

(vii) Start/Stop management node process

Start the Management node process

Nohup Masterha_manager--conf=/etc/masterha/app1.cnf--ignore_last_failover &>/dev/null &

To close the management node process

Masterha_stop–conf=/etc/masterha/app1.cnf

(eight) log view

Tail-f/masterha/app1/manager.log

(ix) Fast switching

Through log filtering, the fault main library is quickly switched from the library.

Grep–i "Change"/masterha/app 1/manager.log # Whichever is the last change master

threeother

MHA Each failover, the process will stop, you need to manually restart the process, if the Lost main library does not switch from the library to re-copy the new main library, you need to change the manager 's configuration file, the process restarts, otherwise.

This article is from the "Floating Phoenix Year" blog, please be sure to keep this source http://liuzhanbin.blog.51cto.com/10060150/1641960

MHA Building--percona Server