Transaction Processing System (TPS)

TPS occurs in a government organization, a service firm, and a manufacturing firm.

Transaction is any business event that generates data worth being captured and stored in a database warehouse.

Properties in TPS:

• Atomicity: the transaction starts and each step must be completed or it will be undone.
• Consistency: data is agreed upon, remains in a consistent state and is accounted for.
• Isolation: treats each transaction separately and keeps data from each transaction separate.
• Durability: the data should continue to exist.

Various types of transaction occur in organizations each day such as a service sold, a payroll check generated, a person hired, and a product manufactured. TPS are inputs for business operations, functional area information systems, and business intelligence systems. Examples of business operations are customer relationship management, knowledge management, and e-commerce.

Problems and issues in TPS:

• Errors resulting from overlapping updates may occur when multiple peoples or application programs can access the database at the same time. Example of error is the results of one of one of the updates are lost.
• Inconsistencies may arise from any failure of any component at any time when a transaction involves more than one computer.
• Eg: error that occurs at some point in an ATM withdrawal
• The customer receive cash although the bank's computer indicates that he did not.
• The customer did not receive cash although the bank's computer indicates that he did.
• TPS must be reversible if error ever occur during transactions. It is also necessary when a purchased item is returned.
• Preserving an audit trail is also important and may be legally required for certain transactions.

Organizations spend a lot of money on expensive mainframe computers to avoid errors and downtime during transaction, record results accurately and securely, and maintain privacy and security. Dependable and reliable computers with the capacity to handle the organization's transaction processing loads are needed for firms to efficiently handle data both in high volumes and during large variations in volume (peak times). Organizations do a process called source data automation which they automate TPS data entry as frequently as they can because of the large volumes of data involved.

Types of TPS:

• Batch processing, as transaction occur, data are collected and placed in groups or batches. Batch processing takes place for transaction that are not able to be easily stored and updated due to the amount and type of data used. Next, the system prepares and process the batches periodically. Example of transaction that uses batch processing is payment using cheque that involves large amount of money.

• Online transaction processing (OLTP), also known as Real time processing, where business transactions are processed online as soon as they occur. Example of OLTP is the Bank ATM, which have user input that requires immediate feedback.