一次MYSQL 服务器性能优化之旅

1.主库性能预警，CPU LOADING 过高
    a.  读写分离：把读操作（查询类，不是写入/修改后即查询数据的）放到DB2节点中进行；  

2.日志文件、数据文件同放在一个磁盘中 /data ，没有分离 。暂时不处理，以后项目可考虑分开。

3.开启了 doublewrite ，在SSD,XFS 磁盘存储数据，可以关闭
mysql> show variables like '%doublewrite%';
    ERROR 2006 (HY000): MySQL server has gone away
    No connection. Trying to reconnect...
    Connection id:    39223899
    Current database: *** NONE ***

    +----------------------------------+----------------+
    | Variable_name                    | Value          |
    +----------------------------------+----------------+
    | innodb_doublewrite               | ON             |
    | innodb_parallel_doublewrite_path | xb_doublewrite |
    +----------------------------------+----------------+
    2 rows in set (0.01 sec)

    mysql> 



4.参数 table_open_cache 调整为 4096

    #配置文件中已设置成 4096 ,但数据库没设置 。此设置可以动态设置，不用重启MYSQL 实例。
        命令：  set global table_open_cache=4096;

    [[email protected] ~]# cat /etc/my.cnf|grep table_open_cache
    table_open_cache = 4096

    mysql> show variables like '%table_open_cache%';
    +----------------------------+-------+
    | Variable_name              | Value |
    +----------------------------+-------+
    | table_open_cache           | 447   |
    | table_open_cache_instances | 16    |
    +----------------------------+-------+
    2 rows in set (0.00 sec)

    mysql> show global status like 'open%tables%';
    +---------------+----------+
    | Variable_name | Value    |
    +---------------+----------+
    | Open_tables   | 432      |
    | Opened_tables | 39578950 |
    +---------------+----------+
    Opened_tables数值非常大，说明cache太小，导致要频繁地open table


5.query_cache_size 查询缓存参数
    5.1配置文件：
        query_cache_size = 0

    5.2 MYSQL 
    mysql> SHOW STATUS LIKE 'Qcache%'; 
    ERROR 2006 (HY000): MySQL server has gone away
    No connection. Trying to reconnect...
    Connection id:    41556755
    Current database: *** NONE ***

    +-------------------------+-------+
    | Variable_name           | Value |
    +-------------------------+-------+
    | Qcache_free_blocks      | 0     |
    | Qcache_free_memory      | 0     |
    | Qcache_hits             | 0     |
    | Qcache_inserts          | 0     |
    | Qcache_lowmem_prunes    | 0     |
    | Qcache_not_cached       | 0     |
    | Qcache_queries_in_cache | 0     |
    | Qcache_total_blocks     | 0     |
    +-------------------------+-------+
    8 rows in set (0.00 sec)

    mysql> 
    mysql> SHOW VARIABLES LIKE 'have_query_cache';
    +------------------+-------+
    | Variable_name    | Value |
    +------------------+-------+
    | have_query_cache | YES   |
    +------------------+-------+
    1 row in set (0.00 sec)

    可配置为 query_cache_size = 128M



6.隔离级别： 现在配置为 可重复读

 transaction_isolation = REPEATABLE-READ

修改语句： set transaction isolation level read committed; //设置读提交级别 可动态修改。


7.从节点延迟问题：

现有参数：
    mysql> show variables like 'slave_parallel_workers';
    +------------------------+-------+
    | Variable_name          | Value |
    +------------------------+-------+
    | slave_parallel_workers | 0     |
    +------------------------+-------+
    1 row in set (0.00 sec)

    mysql> show variables like 'slave_parallel_type';
    +---------------------+----------+
    | Variable_name       | Value    |
    +---------------------+----------+
    | slave_parallel_type | DATABASE |
    +---------------------+----------+
    1 row in set (0.01 sec)


    为了减少延迟现象
        slave-parallel-type=LOGICAL_CLOCK    #基于组提交的并行复制方式 默认值为 DATABASE   
        slave-parallel-workers=4    #并行数量



8.两个从节点可以把一个设置成延迟1天（或12小时）用于做数据备份使用。
    登陆到Slave数据库服务器  ，设置延迟12小时(12*3600)
    mysql>stop slave; 
    mysql> CHANGE MASTER TO MASTER_DELAY = 43200;
    mysql>start slave; 
    mysql>show slave status\G
    查看SQL_Delay的值为600，表示设置成功。