Real-time transaction scheduling in database systems

Abstract

Real-time database systems have transactions with explicit timing constraints, such as ready time, deadlines, and temporal constraints. Meeting the requirements of real-time database systems (RTDBS) requires a balanced and coordinated effort between concurrency control and transaction scheduling. Current research efforts in RTDBS have been focused on scheduling transactions with soft or firm deadlines with serializability as the sole correctness criterion. Efficient resource scheduling algorithms and concurrency control protocols are required to schedule so as to satisfy both timing and data consistency requirements. What we need is a concurrency control algorithm that allows transactions to meet the timing constraints as much as possible without reducing the concurrency level of the system in the absence of any a priori information. In this paper, we present new real-time transaction scheduling algorithms which employ a hybrid approach, i.e., a combination of locking, pessimistic and optimistic methods. Our experimental results indicate that proposed protocols appears to be good, especially under conditions of high transaction load and high data contention.