Concurrency Control in Transactions Processing Systems

Since the birth of shared database systems, conserving consistency of data has been a serious and complicated problem. Concurrency control is the formal terminology given to this bottleneck; when an item is being shared by many users who try to read/update this item simultaneously. During the recent years, this issue has been gaining more importance and attraction due to the dramatic improvement in the fields of networking, and thus the increase use of shared servers by thousands of users at a time. Such systems and have been classified as Transactions Processing Systems (TPS) . This book introduces the concurrency control field of study in TPS in general and then explores a straightforward (can be even applied on single-processor systems) but efficient method (allowing parallelism when possible) to overcome this problem. The proposed technique uses a lock-based approach that guarantees consistency of data and handles a relatively large number of requests in a small period of time. This monograph should light, from a computer scientist perspective, the art and science of sharing a set of data by many end-users without losing the integrity of the data.

Autorentext

Earned his MS and BS in Computer Science from the Lebanese American University and Beirut Arab University respectively. He has been an instructor at the Lebanese International University teaching Database Systems and Algorithms-based Courses. Jad Abbass's research interests include Bio-informatics and Database Systems.