In any computer-based system, data storage and retrieval plays an important role. Data storage involves decision about the encoding of data, assignment of data to meaningful classes and the organization of data in the form of files. On the conceptual level, data is stored in related groupings of records which constitute the files of the system. In a data processing application, two types of files are used:
Master file contains information on a particular instant (historical information). This information has to be updated on the basis of day-to-day transactions. The updating can be done at fixed time intervals or instantaneously (on-line Realtime processing – Batch Processing).
The information generated between the updating cycle has to be stored in what is called as a transaction file. Let us consider an example of an inventory system:
Master file will contain the status of stock position as on a particular time. Each record may contain:
The stock items will be issued and received regularly over the day. The issues and receipts will form the transactions file. It may contain the following data:
As can be seen, the two files have inventory code as a common item. The master file can be updated using the transaction file and generate a new master file. In a realtime system, each transaction is independently processed and the files are updated instantaneously.