MySQL and PostgreSQL methods to grab slow SQL

1. MySQL Medium crawl slow SQL

1. Check the related configuration of slow query

show variables like '%quer%';

• slow_query_log: Indicates whether to turn on the slow query log
• slow_query_log_file: Specify which log file the slow query log is stored in
• long_query_time: Indicates how long the query is recognized as " The slow query ". The default value is ：10s
• log_queries_not_using_indexes: Indicates if it runs sql Indexes are not used , Whether it is also recorded in the slow query log as a slow query statement .
• log_throttle_queries_not_using_indexes： When log_queries_not_using_indexes Set to on when , This parameter limits the number of times a query statement that does not use an index is recorded in the slow query log per minute .

2. Temporarily open / Turn off slow query log

set global slow_query_log = on

set global slow_query_log = off

3. View slow query log
   （1） Just started using slow queries ,vi or cat The command can still be used , You can also find , But as the number of logs grows , You may not be able to query , At this time, you need to use 2
   （2）MySQL Built in mysqldumpslow Command analysis （ The specific use method can be Baidu ）

2. PostgreSQL Grab slow SQL

1. View the current slow query configuration

show log_min_duration_statement;

2. stay postgresql.conf Configure global slow query log in （ millisecond ）

log_min_duration_statement=10000;

3. Slow query log location storage

log_destination = 'csvlog' # After opening this , below log_filename designated log The suffix file will have no content , All in csv In file 
logging_collector = on
log_directory = 'pg_log' # In the current directory pg_log Folder 
log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log'

4. Reload configuration

select pg_reload_conf();

5. test

begin TRANSACTION; 
select pg_sleep(5);
select now() ;
END TRANSACTION;

6. see csv file , Filter duration Field