Robust quality management for differentiated imprecise data services

Several applications, such as web services and e-commerce, are operating in open environments where the workload characteristics, such as the load applied on the system and the worst-case execution times, are inaccurate or even not known in advance. This implies that transactions submitted to a real-time database cannot be subject to exact schedulability analysis given the lack of a priori knowledge of the workload. In this paper we propose an approach, based on feedback control, for managing the quality of service of real-time databases that provide imprecise and differentiated services, given inaccurate workload characteristics. For each service class, the database operator specifies the quality of service requirements by explicitly declaring the precision requirements of the data and the results of the transactions. The performance evaluation shows that our approach provides reliable quality of service even in the face of varying load and inaccurate execution time estimates.