扫二维码与项目经理沟通
我们在微信上24小时期待你的声音
解答本文疑问/技术咨询/运营咨询/技术建议/互联网交流
搭建MySQL的主从复制和读写分离 +--------+ (write) +--------+ | client | +---------------------+| master | +--------+| | +--------+| | | | | | | | +--------+ (read) | (read) +--------+| amoeba |+---------------------|-----------+ (write) | +--------+| | | | | | | | (read)| | | +--------+| | +--------+ | +--------+ | client | +-| slave1 |+---------+----------+| slave2 | +--------+ +--------+ (replication) +--------+ master mysql:172.17.0.4 slave1 mysql:172.17.0.5 slave2 mysql:172.17.0.6 mysql-proxy(amoeba):172.17.0.8 Test host:172.17.0.7 主从复制: master mysql : 安装mysqlserver和mysql yum install mysql-server mysql -y 修改配置文件 vim /etc/my.cnf.d/server.cnf [mysqld] server-id = 1 log-bin = master-bin 启动mysql /etc/init.d/mysql start 修改root密码 mysqladmin -uroot -p password mysql 登陆mysql测试并查看master状态 mysql -uroot -pmysql Welcome to the MariaDB monitor. Commands end with ; or \g. Your MariaDB connection id is 22449 Server version: 5.5.47-MariaDB-log MariaDB Server Copyright (c) 2000, 2015, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MariaDB [(none)]> show master status; +------------------+----------+--------------+------------------+ | File | Position | Binlog_Do_DB | Binlog_Ignore_DB | +------------------+----------+--------------+------------------+ | mysql-bin.000001 | 3648 | | | +------------------+----------+--------------+------------------+ 1 row in set (0.03 sec) 创建主从同步账号 MariaDB [(none)]> grant replication slave on *.* to 'mysqlmaster'@'172.17.0.%' identified by '123456'; slave1 mysql : 安装mysqlserver和mysql yum install mysql-server mysql -y 修改配置文件 vim /etc/my.cnf.d/server.cnf [mysqld] server-id = 2 #id必须唯一 log-bin = slave-bin 启动mysql /etc/init.d/mysql start 修改root密码 mysqladmin -uroot -p password mysql 创建同步文件 MariaDB [(none)]> change master to master_host='172.17.0.4',master_user='mysqlmaster',master_password='123456',master_log_file='mysql-bin.000001',master_log_pos=3648; MariaDB [(none)]> start slave; 查看是否成功,确保下面两项为Yes MariaDB [(none)]> show slave status\G; Slave_IO_Running: Yes Slave_SQL_Running: Yes master mysql : 在master mysql上创建数据查看slave1 mysql是否同步 MariaDB [(none)]> create database new1; Query OK, 1 row affected (0.00 sec) MariaDB [(none)]> create database new2; Query OK, 1 row affected (0.00 sec) MariaDB [(none)]> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | new1 | | new2 | | performance_schema | | test | +--------------------+ 6 rows in set (0.00 sec) slave1 mysql : MariaDB [(none)]> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | new1 | | new2 | | performance_schema | | test | +--------------------+ 6 rows in set (0.00 sec) 我们生产环境中会碰到这种情况:备份主机数据,或者添加一台主机。进行双主机的结构。 另外就是在线上中途添加更多的从机。 而我们知道,从机上设置 slave 时要指定 master_log_file 和 master_log_pos, 即指定binlog文件和偏移值。 这也就是说,从机是可以从任意位置的 binlog 文件中进行数据的同步。比如:我们将 binlog 文件备份到其它某处放置,某天,数据库出问题了,需要对某些数据进行数据恢复,这时候从该文件中进行恢复。 添加一个新的从机,可以有两种方式:从 master 机器复制; 另一种是直接从 slave 复制. mysql-主从结构添加新的slave 两种解决办法 1.copy master master mysql : 锁定数据库 MariaDB [(none)]> flush tables with read lock; Query OK, 0 rows affected (0.04 sec) 查看主机状态,几下file position参数 MariaDB [(none)]> show master status; +------------------+----------+--------------+------------------+ | File | Position | Binlog_Do_DB | Binlog_Ignore_DB | +------------------+----------+--------------+------------------+ | mysql-bin.000001 | 3974 | | | +------------------+----------+--------------+------------------+ 备份所有数据库 mysqldump --all-databases -uroot -pmysql > backup.sql 拷贝到准备新加的slave主机 scp backup.sql root@172.17.0.6:/root slave2 mysql : 安装mysqlserver和mysql yum install mysql-server mysql -y 修改配置文件 vim /etc/my.cnf.d/server.cnf [mysqld] server-id = 3 #id必须唯一 log-bin = slave2-bin 启动mysql /etc/init.d/mysql start 修改root密码 mysqladmin -uroot -p password mysql 导入主服务器scp过来的数据库 mysql -uroot -pmysql < backup.sql 创建同步文件 MariaDB [(none)]> change master to master_host='172.17.0.4',master_user='mysqlmaster',master_password='123456',master_log_file='mysql-bin.000001',master_log_pos=3974; MariaDB [(none)]> start slave; 查看是否成功,确保下面两项为Yes MariaDB [(none)]> show slave status\G; Slave_IO_Running: Yes Slave_SQL_Running: Yes master mysql : 创建数据验证是否添加成功 MariaDB [(none)]> create database new3; Query OK, 1 row affected (0.00 sec) MariaDB [(none)]> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | new1 | | new2 | | new3 | | performance_schema | | test | +--------------------+ 6 rows in set (0.00 sec) slave2 mysql : MariaDB [(none)]> create database new3; Query OK, 1 row affected (0.00 sec) MariaDB [(none)]> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | new1 | | new2 | | new3 | | performance_schema | | test | +--------------------+ 6 rows in set (0.00 sec) 2 copy slave 复制从库要步骤: ==================== 可以看到,从主库复制会有段时间锁表,这段时间会影响主库的使用。如果我们能直接从从库进行复制,就不会对主库产生影响了。但是,从从库复制要保证的是复制过程中从库上的数据不会发生变化,所以要先停掉从库。 1.停止从库: mysql> stop slave; 2.看当前从库的状态。和前面的看主库状态一样。但现在是从从库复制,所以查看从库状态:mysql> show slave status; 记下 Relay_Master_Log_file 和 Exec_Master_Log_Pos, 用处和前面一样. 3.备份从库数据.用 mysqldump 4.在新的从库上还原数据 5.设置新从库的 slave 参数.change master to master_host = '192.168.3.119',master_port = 3306,master_user = 'repl_user',master_password='root',master_log_file='master-bin.000005',master_log_pos=194244; 可以看到,虽然新从库是从从库复制的数据,但实际上 binlog 的 master 还是指向的主库。 另外,这里将 master_log_file 和 master_log_pos 设置成第 2 步中的 Relay_Master_Log_file 和 Exec_Master_Log_Pos start slave; mysql的主从复制+读写分离 +--------+ (write) +--------+ | client | +---------------------+| master | +--------+| | +--------+| | | | | | | | +--------+ (read) | (read) +--------+| amoeba |+---------------------|-----------+ (write) | +--------+| | | | | | | | (read)| | | +--------+| | +--------+ | +--------+ | client | +-| slave1 |+---------+----------+| slave2 | +--------+ +--------+ (replication) +--------+ 读写分离 Amoeba(变形虫)项目,专注 分布式数据库 proxy 开发。座落与Client、DB Server(s)之间。对客户端透明。具有负载均衡、高可用性、sql过滤、读写分离、可路由相关的query到目标数据库、可并发请求多台数据库合并结果。再看下上面的架构图。 因为amoeba是java编写的,所以需要先安装java框架 Amoeba : java安装 下载java wget http://download.oracle.com/otn-pub/java/jdk/7u79-b15/jdk-7u79-linux-x64.tar.gz 创建java目录,并解压 mkdir -p /usr/jdk tar -xzvf jdk-7u79-linux-x64.tar.gz -C /usr/jdk 配置环境变量,在/etc/profile文件最后添加如下配置 vim /etc/profile export JAVA_HOME=/usr/jdk/ export CLASSPATH=${JAVA_HOME}/lib export PATH=${JAVA_HOME}/bin:$PATH 使配置文件生效 source /etc/profile 测试 java -version java version "1.7.0_79" Java(TM) SE Runtime Environment (build 1.7.0_79-b15) Java HotSpot(TM) 64-Bit Server VM (build 24.79-b02, mixed mode) 证明已经安装完成 amoeba安装配置 下载解压 wget http://ufpr.dl.sourceforge.net/project/amoeba/Amoeba%20for%20mysql/3.x/amoeba-mysql-3.0.5-RC-distribution.zip unzip amoeba-mysql-3.0.5-RC-distribution.zip mv amoeba-mysql-3.0.5-RC /usr/local/ 配置 vim amoeba-mysql-3.0.5-RC/conf/amoeba.xml (前段连接文件)....................... ....................... 8066 ///////amoeba监听端口/////////128 64 root ///Amoeba代理用户名/////mysql ///Amoeba代理用户密码/////${amoeba.home}/conf/access_list.conf vim amoeba-mysql-3.0.5-RC/conf/dbServers.xml (后端数据库参数文件) ${amoeba.home}/conf/rule.xml ${amoeba.home}/conf/ruleFunctionMap.xml ${amoeba.home}/conf/functionMap.xml 1500 master ///////默认地址池////////master //////写地址池///////////vipdb //////读地址池////////true ............................ ............................ ${defaultManager} 64 128 3306 /////数据库连接端口///////test ////默认数据库////amoeba ///主从数据库默认连接用户////mysql ////主从数据库默认连接用户密码////+---+ | || | |172.17.0.4 || || |+-------//////定义后端数据库,dbServer可以随意命名,但自己必须清楚哪个是主,那个是从,而且主服务器命名要和amoeba.xml中writePool相对应,从服务器对应下面 |172.17.0.5 |virturl dbServer中的poolNames ip对应各个db///////| | | +---|172.17.0.6 |+---+ +---+ vim amoeba-mysql-3.0.5-RC/jvm.properties (java虚拟机配置) 把原来的这一句JVM_OPTIONS="-server -Xms256m -Xmx1024m -Xss196k -XX:PermSize=16m -XX:MaxPermSize=96m" 改成JVM_OPTIONS="-server -Xms1024m -Xmx1024m -Xss256k -XX:PermSize=16m -XX:MaxPermSize=96m" 在master,slave1,slave2分别授权dbServer.xml中定义的用户名和密码 MariaDB [(none)]> grant all on *.* to 'amoeba'@'%' identified by "mysql"; MariaDB [(none)]> grant all on *.* to 'amoeba'@'localhost' identified by "mysql"; 启动amoeba /usr/local/amoeba-mysql-3.0.5-RC/bin/launcher& 查看进程(如果启动成功,会看到如下进程) ps -ef root 467 337 0 12:51 pts/0 00:00:00 /bin/bash amoeba-mysql-3.0.5-RC/bin/launcher root 472 467 0 12:51 pts/0 00:00:05 /usr/local/jdk1.7.0_67/bin/java -server -Xms1024m -Xmx1024m -Xss256k -XX:PermSize=16m root 473 467 0 12:51 pts/0 00:00:00 tail -f /usr/local/amoeba-mysql-3.0.5-RC/logs/console.log 查看监听端口 netstat -anpl tcp 0 0 :::8066 :::* LISTEN 472/java tcp 0 0 ::ffff:172.17.0.8:8066 ::ffff:172.17.0.7:39978 ESTABLISHED 472/java tcp 0 0 ::ffff:172.17.0.8:46624 ::ffff:172.17.0.5:3306 ESTABLISHED 472/java tcp 0 0 ::ffff:172.17.0.8:41286 ::ffff:172.17.0.4:3306 ESTABLISHED 472/java Test host测试 测试是拿amoeba.xml中定义的用户名密码去测试,千万别被这地方绕进去 [root@7898596a875b ~]# mysql -uroot -p -h272.17.0.8 -P8066 Enter password: Welcome to the MariaDB monitor. Commands end with ; or \g. Your MySQL connection id is 1265630343 Server version: 5.1.45-mysql-amoeba-proxy-3.0.4-BETA MariaDB Server Copyright (c) 2000, 2016, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MySQL [(none)]> show databases; +--------------------+ | Database | +--------------------+ | information_schema | | mysql | | new1 | | new2 | | performance_schema | | test | +--------------------+ 6 rows in set (0.00 sec) 测试读写分离 master mysql : 在master上建表 MariaDB [(none)]> use new1; MariaDB [new1]> create table student( id int(4) not null AUTO_INCREMENT, name char(20) not null, age tinyint(2) NOT NULL default '0', dept varchar(16) default NULL, primary key(id), KEY index_name (name) ); slave1,slave2 : slave1 停掉slave MariaDB [(none)]> stop slave; Query OK, 0 rows affected (0.02 sec) MariaDB [new1]> insert into new1.student(id,name) values(2,'slave'); Query OK, 1 row affected (0.03 sec) MariaDB [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 2 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) slave2 MariaDB [(none)]> stop slave; Query OK, 0 rows affected (0.02 sec) MariaDB [(none)]> insert into new1.student(id,name) values(3,'slave'); Query OK, 1 row affected (0.04 sec) MariaDB [(none)]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 3 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) master上也插入一条数据 MariaDB [new1]> insert into new1.student(id,name) values(1,'slave'); Query OK, 1 row affected (0.02 sec) 查询插入 MariaDB [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 1 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) Test host测试 MySQL [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 3 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) 第一次查询发现只查询到slave的数据,以为另一个slave读写分离没有做成功,再次查询发现 amoeba是轮询着从从数据库池里读取数据 MySQL [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 2 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) 再插入一条数据,发现查询不到插入的,还是只能查询到slave上数据 MySQL [new1]> insert into student(id,name) values(4,'yufyang'); Query OK, 1 row affected (0.02 sec) MySQL [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 3 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.00 sec) MySQL [new1]> select * from new1.student; +----+-------+-----+------+ | id | name | age | dept | +----+-------+-----+------+ | 2 | slave | 0 | NULL | +----+-------+-----+------+ 1 row in set (0.01 sec) master mysql : MariaDB [new1]> select * from new1.student; +----+---------+-----+------+ | id | name | age | dept | +----+---------+-----+------+ | 1 | slave | 0 | NULL | | 4 | yufyang | 0 | NULL | +----+---------+-----+------+ 2 rows in set (0.00 sec) 发现刚才在测试机上通过amoeba插入的数据已经出现在主数据库的表中 开启slave 再次查询slave上的数据 MariaDB [new1]> select * from new1.student; +----+---------+-----+------+ | id | name | age | dept | +----+---------+-----+------+ | 1 | slave | 0 | NULL | | 2 | slave | 0 | NULL | | 4 | yufyang | 0 | NULL | +----+---------+-----+------+ 发现已经同步master,amoeba测试机上的数据了| | |1 | |+-------///第一行的dbserver name命名一定和amoeba.xml的readPool相对应,property name配置从服务器的集合,可以是一个,可以是多个! |slave1,slave2 |
我们在微信上24小时期待你的声音
解答本文疑问/技术咨询/运营咨询/技术建议/互联网交流