MySQL Tutorials - Herong's Tutorial Examples - v4.46, by Herong Yang
Delete Records That Are Not Referenced
This section provides a tutorial example on bulk deletion for records that are not referenced by another table.
There are two options to delete records that are not referenced by another table. For example, deleting all orders that has no order lines can be done by:
1. Using "NOT IN (...)":
DELETE from Orders where ID NOT IT (select OrderID from OrderLines);
2. Using "LEFT JOIN":
DELETE Orders from Orders o LEFT JOIN OrderLines l ON o.ID = l.OrderID where l.CaseID is null;
Let's compare try both options and see their performance difference.
1. Using "NOT IN (...)" to delete records that are not referenced. I see that deleting 6% of 687533 rows takes 1.80 seconds.
mysql> select count(*) from Orders | 687533 | mysql> drop Table Orders_Test; mysql> create Table Orders_Test like Orders; mysql> insert into Orders_Test select * from Orders; mysql> select count(*) from Orders_Test where ID NOT IN (select OrderID from OrderLines); | 46829 | mysql> select 46829/687533; | 0.0681 | mysql> DELETE from Orders_Test where ID NOT IN (select OrderID from OrderLines); Query OK, 46829 rows affected (1.80 sec) mysql> drop Table Orders_Test;
2. Using "LEFT JOIN" to delete records that are not referenced. I see that deleting 6% of 687533 rows takes 2.04 seconds.
mysql> create Table Orders_Test like Orders; mysql> insert into Orders_Test select * from Orders; mysql> select count(*) from Orders_Test o -> LEFT JOIN OrderLines l ON o.ID = l.OrderID where l.OrderID is null | 46829 | mysql> DELETE o from Orders_Test o -> LEFT JOIN OrderLines l ON o.ID = l.OrderID where l.OrderID is null Query OK, 46829 rows affected (2.04 sec) mysql> select count(*) from Orders_Test where ID NOT IN (select CaseID from CaseMolecules); | 0 | mysql> drop Table Orders_Test;
Conclusion: the "LEFT JOIN" option takes slightly more execution time than the "NOT IN" option.
Table of Contents
MySQL Introduction and Installation
Introduction of MySQL Programs
Perl Programs and MySQL Servers
Java Programs and MySQL Servers
Character Strings and Bit Strings
Table Column Types for Different Types of Values
Using DDL to Create Tables and Indexes
Using DML to Insert, Update and Delete Records
Using SELECT to Query Database
Window Functions for Statistical Analysis
Use Index for Better Performance
Transaction Management and Isolation Levels
Defining and Calling Stored Procedures
Variables, Loops and Cursors Used in Stored Procedures
System, User-Defined and Stored Procedure Variables
Storage Engines in MySQL Server
InnoDB Storage Engine - Primary and Secondary Indexes
Performance Tuning and Optimization
General Guidelines on Bulk Changes
Bulk Delete with a Simple Condition
►Delete Records That Are Not Referenced
Reset AUTO_INCREMENT Value on Large Tables
Installing MySQL Server on Linux