What Is a Transaction (Magic xpa 3.x)

What are transactions? Why do we need to use transactions?

Transactions are an integral part of developing data-bound applications. A key to developing applications in a database environment is the ability to optimally use transactions to ensure data integrity.

The word “transaction” is used very often when discussing SQL applications. A transaction is an integral part of a process, contributing to the composition of the whole application, and it can be defined as the execution of a set of logically related data modifications, which must be committed (completed and written to disk) or aborted as a single unit.

This means that either the entire process succeeded or the entire process failed. There is no middle ground. Several operations such as UPDATE, DELETE and INSERT may create a single unit. Only if all the operations are successful, will this logical unit be successful.

A transaction process may be an entire business logic process or a smaller unit that is part of a business logic process.

Transactions can be used to secure Read operations, as opposed to Read/Write operations. A Read transaction ensures that the data read within the transaction is not modified by other users.

The transaction processing technique automatically logs all of the updates of a transaction to a temporary transaction file. The updates in this file are cleared only when the transaction is complete, that is, the updates to all the regular database tables have been completed successfully. If a problem is detected in any of the tables affected by the transaction, the entire transaction is canceled and the database is rolled back, that is, restored to its original state before the transaction occurred. The rollback uses the information stored in the Transaction Log file to restore the project.

In this section, you can find information about the following:

Why We Need Transactions

What About Multi-User Environments?

Transaction Pitfalls

Isolation Levels

Locked Records