Search⌘ K
AI Features
Home
Courses
Introduction to Modern Databases

Transaction Control in SQL

Understand the purpose and use of SQL transaction control commands including START TRANSACTION, COMMIT, ROLLBACK, and SAVEPOINT. This lesson helps you ensure data integrity by grouping multi-step operations into logical units, handle errors gracefully, and maintain consistent data in real-world database scenarios.

We'll cover the following...

Imagine we’re processing a customer’s large order in our OnlineStore database.

The system updates stock quantities, loyalty points, and the order records. But what if, midway through this process, a failure occurs—perhaps a system crash or a data error? Without proper safeguards, we could end up with inconsistent data: reduced stock but no recorded order.

That’s where transaction control in SQL becomes essential.

By the end of this lesson, we will be able to:

  • Understand the purpose of transaction control commands in SQL.

  • Use START TRANSACTION to start a new transaction.

  • Use COMMIT to permanently save the changes of a transaction.

  • Use ROLLBACK to undo the changes of a transaction if an error occurs.

  • Apply these commands to ensure data integrity in multi-step operations.

Transaction control

We have already learned about the ACID properties, which guarantee that transactions are reliable.

Most simple INSERT, UPDATE, or DELETE statements run as implicit, single-statement transactions. The database automatically starts a transaction before the statement runs and commits it if the statement succeeds, or rolls ...