Mysql的主从复制和级联_MySQL

bitsCN.com

Mysql的主从复制和级联

 

记录一下mysql的主主复制和级联复制。 

一、环境

主机:192.25.10.71

主机:192.25.10.73

从机:192.25.10.76

DB:Mysql 5.5.31

OS:CentOS 6.3

二、架构  

 

 

三、实施步骤 

mysql的安装略 

1.master端创建测试DB

[root@proxy1 ~]# mysqlWelcome to the MySQL monitor.  Commands end with ; or /g.Your MySQL connection id is 1959Server version: 5.5.31-log Source distributionCopyright (c) 2000, 2013, Oracle and/or its affiliates. All rights reserved.Oracle is a registered trademark of Oracle Corporation and/or itsaffiliates. Other names may be trademarks of their respectiveowners.Type 'help;' or '/h' for help. Type '/c' to clear the current input statement.mysql> create database db_kenyon;Query OK, 1 row affected (0.03 sec)mysql> show databases;+--------------------+| Database           |+--------------------+| information_schema || db_kenyon          || mysql              || performance_schema || test               |+--------------------+5 rows in set (0.00 sec)mysql> use db_kenyon;Database changedmysql> create table t_kenyon(id int);Query OK, 0 rows affected (0.07 sec)mysql> insert into t_kenyon values(1),(2),(3);Query OK, 3 rows affected (0.01 sec)Records: 3  Duplicates: 0  Warnings: 0mysql> select * from t_kenyon;+------+| id   |+------+|    1 ||    2 ||    3 |+------+3 rows in set (0.00 sec)mysql> show master status;Empty set (0.00 sec)[root@localhost ~]#2.修改配置/etc/my.cnf[mysqld]log-bin=mysql-binserver-id    =1 binlog-do-db=db_kenyoninnodb_data_home_dir = /usr/local/mysql/data/innodb_log_group_home_dir = /usr/local/mysql/data/innodb_buffer_pool_size = 256Minnodb_additional_mem_pool_size = 20Minnodb_log_file_size = 64Minnodb_log_buffer_size = 8Minnodb_flush_log_at_trx_commit = 1innodb_lock_wait_timeout = 503.创建slave端连到master端的用户,重启master[root@localhost ~]# /etc/init.d/mysqld restartShutting down MySQL.. SUCCESS! Starting MySQL............ SUCCESS![root@localhost ~]# mysqlmysql>grant replication slave on *.* to repl@'192.25.10.73' identified by '123456';mysql> flush privileges;[root@localhost ~]#4.备份master端数据,并拷贝至slave端mysql> flush tables with read lock;Query OK, 0 rows affected (0.00 sec)mysql>  show master status;+------------------+----------+---------------+--------------------------------------------------+-------------------+| File             | Position | Binlog_Do_DB  | Binlog_Ignore_DB                                 | Executed_Gtid_Set |+------------------+----------+---------------+--------------------------------------------------+-------------------+| mysql-bin.000002 |      120 | db_kenyon |  |                   |+------------------+----------+---------------+--------------------------------------------------+-------------------+1 row in set (0.00 sec)mysql> 另外一个session:[root@localhost log]# cd /usr/local/mysql/data[root@localhost data]# tar -zcvf backup.tar.gz db_kenyon[root@localhost ]# scp backup.tar.gz root@192.25.10.73:/usr/local/mysql/data/ --回到之前的session,从机搭完以后解锁，保证数据一致mysql> unlock tables;Query OK, 0 rows affected (0.00 sec)5.修改slave端的数据[mysqld]server-id  = 2master-host     =   192.25.10.71   #5.5以上废弃master-user     =   repl            #5.5以上废弃master-password =   123456           #5.5以上废弃master-port     =   3306                  #5.5以上废弃master-connect-retry=60           #重试时间60秒, #5.5以上废弃replicate-do-db=db_kenyon         #需要同步的库log-slave-updates=1               #启用从库日志，这样可以设置链式复制read-only = 1   #0表示可读写，1表示只读,但是只针对普通用户，超级用户和同步用户不受此限制6.重启slave，并指定主机检查mysql> change master to master_host='192.25.10.71',master_user='repl',master_password='123456',master_log_file='mysql-bin.000002',master_log_pos=120mysql> show slave status/G*************************** 1. row ***************************               Slave_IO_State:                  Master_Host: 192.25.10.71                  Master_User: repl                  Master_Port: 3306                Connect_Retry: 60              Master_Log_File: mysql-bin.000002          Read_Master_Log_Pos: 120               Relay_Log_File: localhost-relay-bin.000001                Relay_Log_Pos: 4        Relay_Master_Log_File: mysql-bin.000002             Slave_IO_Running: No            Slave_SQL_Running: No              Replicate_Do_DB: db_kenyon          Replicate_Ignore_DB:           Replicate_Do_Table:       Replicate_Ignore_Table:      Replicate_Wild_Do_Table:  Replicate_Wild_Ignore_Table:                   Last_Errno: 0                   Last_Error:                 Skip_Counter: 0          Exec_Master_Log_Pos: 120              Relay_Log_Space: 120              Until_Condition: None               Until_Log_File:                Until_Log_Pos: 0           Master_SSL_Allowed: No           Master_SSL_CA_File:           Master_SSL_CA_Path:              Master_SSL_Cert:            Master_SSL_Cipher:               Master_SSL_Key:        Seconds_Behind_Master: NULLMaster_SSL_Verify_Server_Cert: No                Last_IO_Errno: 0                Last_IO_Error:               Last_SQL_Errno: 0               Last_SQL_Error:  Replicate_Ignore_Server_Ids:             Master_Server_Id: 0                  Master_UUID:             Master_Info_File: /usr/local/mysql/data/master.info                    SQL_Delay: 0          SQL_Remaining_Delay: NULL      Slave_SQL_Running_State:           Master_Retry_Count: 86400                  Master_Bind:      Last_IO_Error_Timestamp:     Last_SQL_Error_Timestamp:               Master_SSL_Crl:           Master_SSL_Crlpath:           Retrieved_Gtid_Set:            Executed_Gtid_Set:                Auto_Position: 01 row in set (0.00 sec)

调整网络和mysql参数等，使两个参数：Slave_IO_Running， Slave_SQL_Running的状态为YES

7.级联过程

其他配置如上，注意级联参数log-slave-updates开启,该参数在my.cnf里值应为1，或者

mysql> show variables like 'log_slave_%';+-------------------+-------+| Variable_name     | Value |+-------------------+-------+| log_slave_updates | ON   |+-------------------+-------+1 row in set (0.00 sec)

8.主主的过程就是主从的基础上再以从机为主机，反向操作一遍即可。简单测试，在71上建表，73、76上都能看到该表即成功。

四、级联的性能损耗测试

使用mysqlslap

[root@proxy1 bin]# ./mysqlslap   --concurrency=2  --iterations=5 --number-int-cols=10 --number-char-cols=10 -a --auto-generate-sql-guid-primary --number-of-queries=10000 --auto-generate-sql-load-type=writeBenchmark        Average number of seconds to run all queries: 5.324 seconds        Minimum number of seconds to run all queries: 4.106 seconds        Maximum number of seconds to run all queries: 6.113 seconds        Number of clients running queries: 2        Average number of queries per client: 5000

 

每秒处理能力1878

改变concurrency,得到如下结果:

 

并发数 平均时间s(后者开启级联) 每秒处理 损耗

2 5.324 1878 0.6s  11%

2 5.912 1691 0.6s  11%

10 2.521 3966 0.3s  12%

10 2.828 3536 0.3s  12%

50 2.161 4627 0.2s  7%

50 2.331 4290 0.2s  7%

100 2.355 4246 0.3s  11%

100 2.618 3819 0.3s  11%

bitsCN.com