MySQL parsing under high concurrency

Work in the case of SQL optimization, often have such a problem, SQL has all the use of the index, the requested amount of data is small, the total amount of data is not large, there are still many slow query appearance (we stipulate is >20ms). This time we should pay attention to our concurrency, transaction locks, the situation.

For the game, DB has a lot of inserts, update can be said that many players will communicate with DB.

This article temporarily ignores IO utilization in the OS, network card traffic, CPU changes, how to view the MySQL part of the parameters

To view the number of transactions per second:

Show global status like ' Com_commit ';

Show global status like ' Com_rollback ';

Tps= (Com_commit+com_rollback)/seconds

To view the number of queries per second:

Mysqladmin status Remove QPS

Query SQL Execution Status:

Show Processlist; Intercepting SQL, executing State,

To view the number of connections:

Show status like ' threads% ';

threads_connected number of threads already open

Threads_running the active thread

This column more highlights: http://www.bianceng.cn/database/MySQL/

For MySQL high concurrency optimization experience can refer to: http://www.gxldan.com/archives/203.html

High performance MySQL third edition download link: http://vdisk.weibo.com/s/hhS9w

Production Environment Maintenance Reference: Http://www.mysqlops.com/2011/09/02/mysql-fault-manual.html

The above status script will be released later!