Predictability and consistency in real-time transaction processing

A real-time database system (RTDBS) can be de ned as a database system where transactions are associated with real-time constraints typically in the form of deadlines. The system must process transactions so as to both meet the deadlines and maintain the data consistency. Previous research e ort in this eld has been focused on scheduling transactions with soft or rm deadlines under the conventional transaction model and database system architecture which cannot support predictable real-time transaction processing. In this thesis, we provide a framework to realize predictable real-time transaction processing, satisfying both timing and consistency constraints of a real-time database system. First, we classify data objects and transactions found in typical real-time database applications, considering their distinct characteristics and requirements. Each type of real-time data objects has its own correctness criteria, di erent from the conventional one. Real-time transactions are categorized, according to their timing constraints, arrival patterns, data access patterns, availability of data and run-time requirements, and accessed data type. Our model is a superset of conventional models; it includes both hard and soft real-time transactions, and supports the temporal consistency as well as the logical consistency of the database. Second, we develop an integrated transaction processing scheme that extends a xedpriority-based task scheduling framework for mixed task sets into a transaction processing environment, combining it with best-e ort real-time transaction scheduling algorithms. This scheme provides predictability for a RTDBS in the sense that under our transaction processing scheme it is guaranteed that every application in the system will achieve its own performance goals. Along with the transaction processing scheme, a temporal consistency enforcement scheme called Static Temporal Consistency Enforcement (STCE) is introduced.