Incentive-Compatible Priority Pricing and Transfer Analysis in Database Services

A primary concern of physical database design has been efficient retrieval and update of a record because predictable performance of a DBMS is indispensable to time-critical missions. To maintain such phenomenal performance, database manages often spends more than or as much as the goal of an organization can warrant. The motivation of this research stems from the fact that even predictable performance of a physical database can be hampered by stochastic query processing time, physical configurations of a database, and random arrival processes of queries. They all together affect the overall performance of a DBMS. In particular, if there are queuing delays due to limited capacity or during on-peak congestion, this paper suggest to prioritize database services. A surprising finding of this paper is that such a transition from a non-priority system to a corresponding priority-based system can be Pareto-improving in the sense that no users in the system will be worse off after the transition. Thus prioritizing database services can be a viable option for efficient database management.