MaxScale Binlog Server实践

Arce

MaxScale Binlog Server实践
简介
　　Part1:写在最前
　　在之前的博文中有说到MaxScale，作为中间件，配合MHA使用或者主从使用可实现读写分离和负载均衡，今天简单介绍下MaxScale作为Binlog Server来减少主从延迟的问题；MySQL的主从架构中，链式拓扑的架构比较容易出现主从延迟的问题。本文着重介绍MaxScale作为Binlog Server是如何降低主从延迟的。
　　

　　MaxScale配合MHA请移步至：
　　http://suifu.blog.51cto.com/9167728/1869520
　　

　　Part2:本文环境
　　HE1:192.168.1.248 slave
　　HE3:192.168.1.250 master
　　HE4:192.168.1.251 maxscale
　　

　　

架构演示

　　

效果对比

实战

　　Part1:安装maxscale
　　[root@HE4 ~]# yum -y install maxscale-2.0.1-2.centos.6.x86_64.rpm
　　[root@HE4 ~]# mkdir -p /data/binlog
　　[root@HE4 ~]# useradd maxscale
　　[root@HE4 ~]# chown -R maxscale. /data/binlog
　　[root@HE4 ~]# cat /etc/maxscale.cnf

[maxscale]
threads=1
##根据CPU核数设置
[Replication]
type=service
router=binlogrouter
user=mysync
passwd=MANAGER
# 使用主库上的repl复制账号
# 权限：
#   GRANT REPLICATION SLAVE,REPLICATION CLIENT ON *.* TO 'repl'@'%' IDENTIFIED BY   'repl';
router_options=server_id=1251,heartbeat=30,binlogdir=/data/binlog,transaction_safety=1,mariadb10-compatibility=1,send_slave_heartbeat=1
# server_id设置maxscale的，记得不能与主和从库重复，要唯一
#   heartbeat=30秒，意思为当maxscale在30秒内没有接收到主库推送的binlog日志，发送心跳检查
#   binlogdir设置接收binlog的存放路径，目录属性chown -R maxscale.maxscale   /data/binlog
# transaction_safety=1此参数用于启用binlog日志中的不完整事务检测。   当MariaDB MaxScale启动时，如果当前binlog文件已损坏或找到不完整的事务，则可能会出现错误消息。   在正常工作期间，binlog事件不会分配到从库，直到事务已经提交。 默认值为off，设置transaction_safety = on以启用不完全事务检测。
#   send_slave_heartbeat=1开启心跳检查
[Replication   Listener]
type=listener
service=Replication
protocol=MySQLClient
port=5308
# 后端的从库CHANGE   MASTER TO这个端口，默认5308
[CLI]
type=service
router=cli
[CLI Listener]
type=listener
service=CLI
protocol=maxscaled
port=6603

　　

　　Part2:启动Maxscale
　　[root@HE4 ~]# /etc/init.d/maxscale start
　　Starting MaxScale: maxscale (pid 16680) is running...      [  OK  ]
　　[root@HE4 ~]# /etc/init.d/maxscale status
　　Checking MaxScale status: MaxScale (pid  16680) is running.[  OK  ]
　　

　　Part3:从库配置

[root@HE1 ~]# mysql -umysync -pMANAGER -h192.168.1.251 -P5308
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MySQL connection id is 3196
Server version: 10.0.0 2.0.1-maxscale
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)]>  CHANGE MASTER TO MASTER_HOST='192.168.1.250',MASTER_USER='mysync',MASTER_PASSWORD='MANAGER',MASTER_PORT=3306,MASTER_LOG_FILE='mysql-bin.000005',MASTER_LOG_POS=20;
ERROR 1234 (42000): Can not set MASTER_LOG_POS to 20: Permitted binlog pos is 4. Specified master_log_file=mysql-bin.000005
MySQL [(none)]>  CHANGE MASTER TO MASTER_HOST='192.168.1.250',MASTER_USER='mysync',MASTER_PASSWORD='MANAGER',MASTER_PORT=3306,MASTER_LOG_FILE='mysql-bin.000005',MASTER_LOG_POS=4;
MySQL [(none)]> start slave;
Query OK, 0 rows affected (0.00 sec)

　　这里可以看出，Maxscale binlog server只能从位置4开始配置
　　

　　配置好后，在/data/binlog下生成的binlog文件
　　[root@HE4 ~]# cd /data/binlog/
　　[root@HE4 binlog]# ls
　　cache  master.ini  mysql-bin.000003
　　

　　Part4:主库配置

[root@HE3 ~]# mysql -uroot -p
Enter password: 
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MariaDB connection id is 7
Server version: 10.1.16-MariaDB 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.
MariaDB [(none)]> show master status;
+------------------+----------+--------------+------------------+
| File       | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000005 |    652 |        |          |
+------------------+----------+--------------+------------------+
1 row in set (0.00 sec)
MariaDB [(none)]>  grant replication client,replication slave on *.* to 'mysync'@'192.168.1.%' identified by 'MANAGER';
MariaDB [(none)]>flush privileges;

　　

　　

　　Part5:主从配置

从库指向binlogserver
[root@HE1 ~]# mysql -uroot -pMANAGER
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MariaDB connection id is 5
Server version: 10.1.16-MariaDB 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.
MariaDB [(none)]> CHANGE MASTER TO MASTER_HOST='192.168.1.251',MASTER_USER='mysync',MASTER_PASSWORD='MANAGER',MASTER_PORT=5308,MASTER_LOG_FILE='mysql-bin.000005',MASTER_LOG_POS=652;
Query OK, 0 rows affected (0.02 sec)
MariaDB [(none)]> start slave;
Query OK, 0 rows affected (0.00 sec)
MariaDB [(none)]> show slave status\G
*************************** 1. row ***************************
         Slave_IO_State: Waiting for master to send event
          Master_Host: 192.168.1.251
          Master_User: mysync
          Master_Port: 5308
        Connect_Retry: 60
        Master_Log_File: mysql-bin.000005
      Read_Master_Log_Pos: 652
         Relay_Log_File: mysql-relay-bin.000002
        Relay_Log_Pos: 537
    Relay_Master_Log_File: mysql-bin.000005
       Slave_IO_Running: Yes
      Slave_SQL_Running: Yes
        Replicate_Do_DB: 
      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: 652
        Relay_Log_Space: 835
        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: 0
Master_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: 1250
         Master_SSL_Crl: 
       Master_SSL_Crlpath: 
           Using_Gtid: No
          Gtid_IO_Pos: 
    Replicate_Do_Domain_Ids: 
  Replicate_Ignore_Domain_Ids: 
        Parallel_Mode: conservative
1 row in set (0.00 sec)

　　

　　

——总结——
生产环境中，大多采用的是一主多从架构，例如星状拓扑和链式拓扑，星状拓扑在从库过多的情况下，会增加主库的io压力，而链式拓扑虽然缓解了主库的网络IO压力，但其缺点是：二级Slave得到最新的数据，需要再经过一层的复制才到达，期间的延迟比一主多从架构要大。而采用maxscale binlog server则避免了这类问题。由于笔者的水平有限，编写时间也很仓促，文中难免会出现一些错误或者不准确的地方，不妥之处恳请读者批评指正。