MySQL highly available cluster – MHA

What is? MHA

MHA(Master High Availability) It's automatic master Fail over and slave Upgraded packages . It's based on standards MySQL Copy ( asynchronous / Half a synchronous ). MHA There are two parts :MHA Manager( The management node ) and MHA Node( Data nodes ). MHA Manager It can be deployed on a single machine to manage multiple master-slave colony , It can also be deployed in one slave On .MHA Manager Detecting the cluster's node node , If I found master In case of failure , It can automatically send slave Upgrade to a new master, And then all the others slave Point to the new master. The entire fail over process is transparent to the application . MHA node Run on each MySQL Server (master/slave/manager), It has parsing and cleaning through monitoring logs Function to speed up the failover process . As a precondition , It should be configured first MySQL Copy , And set up SSH Public key password free login . Let's say CentOS For example , It's better to install first EPEL(http://fedoraproject.org/wiki/EPEL), Otherwise YUM Some packages may not be found .MHA from Node and Manager form ,Node Run on every one MySQL Server , in other words , Whether it's MySQL master server , still MySQL From the server , All have to be installed Node, and Manager It usually runs on a separate server , But if hardware resources are tight , You can also use one MySQL From the server Part-time job Manager Role .

MHA Working principle and tool kit

working principle

From the crash of master Save binary log events (binlog events) - Identify those with the latest updates slave Apply differential relay logs (relay log) To the others slave Application from master Saved binary log events (binlog events) - Upgrade one slave As new master Make others slave Connect new master replicate

Manager Tools

masterha_check_ssh : Check MHA Of SSH To configure masterha_check_repl : Check MySQL Copy masterha_manager : start-up MHA masterha_check_status : Detect current MHA Running state masterha_master_monitor : monitoring master Is it down? masterha_master_switch : Control failover ( Automatic or manual ) masterha_conf_host : Add or remove configured server Information

Node Tools ( These tools are usually made of MHA Manager The script triggers , No need to operate by hand )

save_binary_logs : Save and copy master Binary log apply_diff_relay_logs : Identify different relay log events and apply them to other slave filter_mysqlbinlog : Remove unnecessary ROLLBACK event (MHA This tool is no longer used ) purge_relay_logs : Clear relay logs ( It won't block SQL Threads )

MHA advantage

a、 Automatic failover fast , Seconds to complete the failover b、 Can be combined with semi synchronous replication , Ensure the consistency of master-slave data c、manager Nodes can manage multiple MHA The cluster system d、 At run time ,manager The node just sends periodically ICMP package , Very low impact on performance e、 as long as mysql Replication technology support engine ,MHA It supports f.MHA Enhanced data security g.MHA Data loss is less likely to occur during failover

---

Deploy Mysql One master and two slaves （ Do a good job of master-slave can be ignored ）

Deploy MYSQL A master from

Environmental preparation

IP	edition	service
192.168.1.40	CentOS Linux release 7.6.1810 (Core)	MHA-manager
192.168.1.10(master)	CentOS Linux release 7.6.1810 (Core)	mysql( Lord )、MHA-node
192.168.1.20(slave1)	CentOS Linux release 7.6.1810 (Core)	mysql( from )、MHA-node
192.168.1.30(slave2)	CentOS Linux release 7.6.1810 (Core)	mysql( from )、MHA-node

This experiment centos Minimize installation Four hosts install some commands And shut down the firewall

systemctl stop firewalld && yum -y install net-tools vim wget git make cmake gcc gcc-c++

Copy

Modify the host name and log in without secret

master operation

[[email protected] ~]# vim /etc/hosts
# add to 
192.168.1.10 master
192.168.1.20 slave1
192.168.1.30 slave2
[[email protected] ~]# ssh-keygen
Generating public/private rsa key pair.
Enter file in which to save the key (/root/.ssh/id_rsa): 
Created directory '/root/.ssh'.
Enter passphrase (empty for no passphrase): 
Enter same passphrase again: 
Your identification has been saved in /root/.ssh/id_rsa.
Your public key has been saved in /root/.ssh/id_rsa.pub.
The key fingerprint is:
SHA256:lU9My4J5QGSrEV8wvMfN61Fxu9jIyCX6+WdxjYTykp8 [email protected]
The key's randomart image is:
+---[RSA 2048]----+
|      .+B.. .    |
|       +.B = .. .|
|      . =o=o=. o.|
|       o.o+=+.o. |
|      . S.o+*++.o|
|         .oo++.+o|
|          .+.o  o|
|           oE  o |
|            ..o  |
+----[SHA256]-----+
[[email protected] ~]# ssh-copy-id -i slave1
/usr/bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/root/.ssh/id_rsa.pub"
The authenticity of host 'slave1 (192.168.1.20)' can't be established.
ECDSA key fingerprint is SHA256:dnnDcAA2qVnA31i7mtr9LYJmH2veu2+r4t+19qUSqqw.
ECDSA key fingerprint is MD5:0e:f3:c1:3c:dc:5f:12:66:ae:c9:01:51:66:db:bb:02.
Are you sure you want to continue connecting (yes/no)? yes
/usr/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed
/usr/bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys
[email protected]'s password: 

Number of key(s) added: 1

Now try logging into the machine, with:   "ssh 'slave1'"
and check to make sure that only the key(s) you wanted were added.
[[email protected] ~]# ssh-copy-id -i slave2
/usr/bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/root/.ssh/id_rsa.pub"
The authenticity of host 'slave2 (192.168.1.30)' can't be established.
ECDSA key fingerprint is SHA256:xltTmad7PHB4XORvNjSpTaA4hdjQmix+o3VHfYgwuRE.
ECDSA key fingerprint is MD5:52:f6:2c:cc:d4:75:ce:06:c6:c0:1e:18:7c:4e:2c:73.
Are you sure you want to continue connecting (yes/no)? yes
/usr/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed
/usr/bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys
[email protected]'s password: 

Number of key(s) added: 1

Now try logging into the machine, with:   "ssh 'slave2'"
and check to make sure that only the key(s) you wanted were added.

[[email protected] ~]# hostname master
[[email protected] ~]# bash
[[email protected] ~]# 
[[email protected] ~]# scp /etc/hosts slave1:/etc/hosts
hosts                                                    100%  218   174.0KB/s   00:00    
[[email protected] ~]# scp /etc/hosts slave2:/etc/hosts
hosts                                                    100%  218   117.7KB/s   00:00    

Copy

slave1 operation

[[email protected] ~]# hostname slave1
[[email protected] ~]# bash
[[email protected] ~]# ssh-copy-id -i master
[[email protected] ~]# ssh-copy-id -i slave2

Copy

slave2 operation

[[email protected] ~]# hostname slave2
[[email protected] ~]# bash
[[email protected] ~]# ssh-copy-id -i master
[[email protected] ~]# ssh-copy-id -i slave1

Copy

Four hosts Turn off the firewall / Sandbox

systemctl stop firewalld && setenforce 0

Copy

Install first MYSQL 1. install MySQL Of yum Source

wget http://dev.mysql.com/get/mysql57-community-release-el7-11.noarch.rpm

Copy

yum -y install mysql57-community-release-el7-11.noarch.rpm

Copy

2. installation is complete , Check whether the installation is successful

yum repolist enabled | grep mysql.*

Copy

3. It can be installed mysql 了

yum install mysql-community-server

Copy

4. After the installation is complete, you can start mysql 了

systemctl start mysqld.service

Copy

5. Get random password

grep 'temporary password' /var/log/mysqld.log

Copy

6. If you want to set your own password easily , It comes with a password checking mechanism that doesn't allow you to set it up easily sql Enter Settings

mysql -uroot -p
password： The random password just obtained 
 Changing the default password is complicated and difficult 
mysql>set global validate_password_policy=0;# Set the password check policy to 0
mysql>set global validate_password_length=1;# Set the password length to 1

Copy

7. Change the database password

set password for  user name  @localhost = password(' New password ');

Copy

8. thus MySQL Installation completed

Modify the configuration file for master-slave replication

To configure Master Main library machine

Turn on binlog Set up mysql The unique number of (mysql5.7 That is, new parameters are added in the later version ), And on binlog journal Input vi /etc/my.cnf Go to configuration file , Press Insert Key to enter edit mode , Add the following parameters server-id=1 ( Unique identification of the machine ) log-bin=/var/lib/mysql/mysql-bin ( Turn on binlog)

vim /etc/my.cnf

Copy

Conduct mysql restart

systemctl restart mysqld

Copy

Get into mysql Catalog view binlog Open successfully or not

Get into MySQL Create users and authorize

mysql> grant replication slave on *.* to 'slave'@'192.168.1.%' identified by '123.com';
Query OK, 0 rows affected, 1 warning (0.00 sec)

Copy

To configure Slave From the library machine

slave1

Copy

slave2

Copy

restart MySQL service

systemctl restart mysqld

Copy

Check the main server status

mysql> show master status;
+-------------------+----------+--------------+------------------+-------------------+
| File              | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+-------------------+----------+--------------+------------------+-------------------+
| master-bin.000006 |     1049 |              |                  |                   |
+-------------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)

Copy

Configure the slave database to synchronize with the master database slave1、salve2

mysql> CHANGE MASTER TO MASTER_HOST='192.168.1.10',
    -> MASTER_USER='slave',
    -> MASTER_PASSWORD='123.com',
    -> MASTER_LOG_FILE='master-bin.000006',
    -> MASTER_LOG_POS=1049;
Query OK, 0 rows affected, 2 warnings (0.01 sec)

Copy

Start the synchronization

start slave;

Copy

see slave state , These two items must be yes

show slave status\G;

Copy

Test for synchronization ：

The main database is used to create the database ;

mysql> create database test;
Query OK, 1 row affected (0.00 sec)
mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| test               |
+--------------------+
5 rows in set (0.00 sec)

Copy

see slave1、slave2 Slave Library ;

mysql> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mysql              |
| performance_schema |
| sys                |
| test               |
+--------------------+
5 rows in set (0.00 sec)

Copy

One master and two slaves are successfully set up ！！！

---

Deploy MHA

The required MHA Install package Links

 link ：https://pan.baidu.com/s/1QOgJrkHItlCjwMX60h6s4A 
 Extraction code ：e9bi 
 Copy this content after opening baidu network disk mobile phone App, It is more convenient to operate 

Copy

First step

build MySQL Master-slave （ Do well and skip ）

One thing to pay attention to ：mysql The authorized users of master-slave replication also need all three , Because once one of the main machines goes down , There is one that will replace the master , So you still need a user who can copy binary logs , The three user names must be the same

MySQL Master my.cnf The configuration file

server-id=1
log-bin=master-bin
log-slave-updates=true
relay-log = relay-log-bin

Copy

MySQL From my.cnf The configuration file

Why does the configuration file of the slave have the same as that of the master log-bin, The same reason for authorization , If after the primary outage , From to main , You also need binary log files

a server-id by 2, a server-id by 3

server-id=2
log-bin=master-bin
log-slave-updates=true
relay-log=relay-log-bin
relay-log-index=slave-relay-bin.index
read-only=1	#  Start read-only properties 
relay-log-purge=0	#  High availability for switching , When ascending from master , Keep previous relay logs 

Copy

Database authorization mha user , This user will be used as the management side mysql Database user used when

Three databases log in to the database client respectively , Execute the following command to authorize

grant all on *.* to 'mha'@'192.168.1.%' identified by '123.com';

Copy

Then authorize the master and slave to copy users ( All three databases do )

grant replication slave on *.* to 'slave'@'192.168.1.%' identified by '123.com';

Copy

see MySQL Lord （192.168.1.10） The offset , Used to copy and use from the server

mysql> show master status;
+-------------------+----------+--------------+------------------+-------------------+
| File              | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+-------------------+----------+--------------+------------------+-------------------+
| master-bin.000001 |      1021 |              |                  |                   |
+-------------------+----------+--------------+------------------+-------------------+
1 row in set (0.01 sec)

Copy

It's on two MySQL（192.168.1.20/192.168.1.30） From the following

change master to master_host='192.168.1.10',master_user='slave',master_password='123.comm',master_log_file='master-bin.000001',master_log_pos=1021;
start slave;
show slave status\G  #  Success occurs when the following occurs 
             Slave_IO_Running: Yes
            Slave_SQL_Running: Yes

Copy

The second step

For the convenience of management , Do password free login between the four servers

MHA The management end 192.168.1.40 node1、node2、node3 Again （ Have done something that can be ignored ）

ssh-keygen
ssh-copy-id -i 192.168.1.10
ssh-copy-id -i 192.168.1.20
ssh-copy-id -i 192.168.1.30

Copy

The third step

Install managed nodes on each server

The operation of the four machines is consistent

[[email protected] ~]# rpm -ivh epel-release-latest-7.noarch.rpm

Copy

Use the Internet yum Installation dependencies , You need to use the network yum

yum -y install perl-DBD-mysql perl-DBI

Copy

[[email protected] ~]# rpm -ivh mha4mysql-node-0.56-0.el6.noarch.rpm

Copy

Step four # Please delete the following comments when copying the following code ( So as not to report a mistake ) stay MHA The management end （192.168.1.40） Install Management Pack

In the absence of mysql Install the following packages on the host of

yum -y install perl-Config-Tiny perl-Log-Dispatch perl-Parallel-ForkManager
rpm -ivh mha4mysql-manager-0.56-0.el6.noarch.rpm

Copy

Writing configuration files

[[email protected] ~]# vim /etc/masterha_default.cnf
[server default]    #  Global configuration 
user=mha    #  The management user authorized by the database to the management node 
password=123.com  #  Password of the management user authorized by the database to the management node 
repl_user=slave		#  Users used by master-slave replication 
repl_password=123.com   #  The user password used by master-slave replication 
ssh_user=root    #  Manage the... Used between nodes ssh user 
master_binlog_dir=/var/lib/mysql  # mysql The directory where the binary logs are stored 
remote_workdir=/data/login	#  When you get mysql The primary binary log is stored in this directory of the local computer 
ping_interval=2  #  Manage nodes and mysql Main communication time （ second ）, Over time , Think mysql The main outage 
shutdown_script=""   #  When mysql Scripts executed when the main machine goes down , Can be null 

Copy

Write the configuration file of the management node

[[email protected] ~]# mkdir /etc/mha
[[email protected] ~]# vim /etc/mha/app1.cnf
[server default]  #  Global configuration 
manager_workdir=/var/lib/manager   #  working directory 
manager_log=/var/log/manager/manager.log	#  Log directory ,manager Need to create 
[server1]   #  Node configuration 
hostname=192.168.1.10
port=3306
ssh_port=22
 
[server2]
hostname=192.168.1.20
port=3306
ssh_port=22
candidate_master=1  #  Express mysql After the main outage , Yes server2 To replace the master 
 
[server3]
hostname=192.168.1.30
port=3306
ssh_port=22

Copy

Create directory

mkdir /var/log/manager

Copy

test manager And database nodes ssh Whether the communication is normal

[[email protected] ~]# masterha_check_ssh --global-conf=/etc/masterha_default.cnf --conf=/etc/mha/app1.cnf
#  Finally, it shows ok

Copy

Check whether the master-slave replication between nodes is normal

[[email protected] ~]# masterha_check_repl --global-conf=/etc/masterha_default.cnf --conf=/etc/mha/app1.cnf
#  Finally, it shows ok

Copy

start-up MHA The management end , In a monitored state

[[email protected] ~]# masterha_manager --global-conf=/etc/masterha_default.cnf --conf=/etc/mha/app1.cnf
#  Blocked state 

Copy

verification ：

MySQL Lord 192.168.1.10

systemctl stop mysqld

Copy

Will the Lord mysql After the service simulates the failure , The monitoring management side will use mha The user starts electing another master , To succeed ,down The fallen Master goes online again , It's independent mysql

Create a library on the slave that becomes the master library to verify , Whether others are copied from the library to

Friendship connection

Publisher ： Full stack programmer stack length , Reprint please indicate the source ：https://javaforall.cn/101934.html Link to the original text ：https://javaforall.cn