Teach you to locate online MySQL slow query problem hand by hand, package teaching package meeting

1. The role of slow query log

The slow query log is not enabled by default. It is recommended to manually enable it to help us locate online problems.

SQL whose execution time exceeds the threshold will be written to the slow query log, which can help us record SQL statements that execute too long, locate online slow SQL problems, and facilitate SQL performance tuning.

2. Configuration of slow query log

2.1 Check whether the slow query log is enabled

show variables like 'slow_query_log';Copy code

The default is OFF, if it is not turned on, it can be turned on manually.

2.2 Enable slow query log

One way is to use the MySQL command to open:

set global slow_query_log=1;Copy code

Another method is to modify the MySQL configuration file, restart the MySQL service, and enable it.

Modify the configuration file my.cnf and add the following command

slow_query_log = ON

2.3 Set the threshold of slow query log

The default threshold for slow query logs is 10, in seconds.

For an online service, 10 seconds is too long, we can modify it manually.

One is to modify it through MySQL commands, such as modifying it to 1 second:

set long_query_time=1;Copy code

Another method is to modify the MySQL configuration file, restart the MySQL service, and enable it.

Modify the configuration file my.cnf and add the following command

long_query_time = 1

2.4 Modify slow query log location

Use the MySQL command to view the slow query log location:

show variables like '%slow_query_log_file%';Copy code

If you want to modify the slow query log location, you can modify the MySQL configuration file, restart the MySQL service, and enable it.

Modify the configuration file my.cnf and add the following command

slow_query_log_file = /usr/local/mysql/data/localhost_slow.log

2.5 Record more slow query SQL

By default, management statements are not recorded in the slow query log. Management statements include ALTER TABLE, ANALYZE TABLE, CHECK TABLE, CREATE INDEX, DROP INDEX, OPTIMIZE TABLE, and REPAIR TABLE.

Management statements are also very important. If you want to be logged, you can modify them through MySQL commands:

set global log_slow_admin_statements=ON;Copy code

By default, statements that do not use an index will not be logged.

Enough to deceive people!If you are not careful, you will fall into the pit. If you don't record the statements that do not use the index, why do you need to query the log slowly?

If you want to record a statement that does not use an index, you can modify it with the command:

set global log_queries_not_using_indexes=ON;Copy code

3. Use of slow query log

Manually create a slow SQL to test the effect. There are 1 million table data in the user table:

select * from user;Copy code

Then take a look at the contents of the slow query log file:

cat /usr/local/mysql/data/localhost_slow.logCopy code

SQL statements and execution times are logged.

4. Analyze slow query logs

Sometimes there are many slow query logs, and it is not very convenient to view them manually. Fortunately, MySQL provides a tool for analyzing slow query logs mysqldumpslow .

 Common parameters are-s: Indicates how to sort:c: number of visitsl: lock timer: return recordt: query timeal: average lock timear: Average number of records returnedat: average query time-t: Return the number of previous data;Copy code

4.1 The 10 SQLs with the most returned results:

mysqldumpslow -s r -t 10 /usr/local/mysql/data/localhost_slow.log

4.2 The 10 SQLs with the longest query time:

mysqldumpslow -s t -t 10 /usr/local/mysql/data/localhost_slow.log