We propose to develop a framework which provides the ability to apply complex event processing in realtime domains, thus allowing an easier process of developing and maintaining specific solutions for real-time event-based systems, while upholding the real time requirements of the system. Specifically, we propose to develop a framework that includes an integrated development environment for defining rules, and, given a set of rules, generates code for a complex event processing application for which it is able to determine time bounds on the response of this application to a set of supported events. In particular, the tool helps determine a time bound for the execution time of the code corresponding to each rule. Many Service Oriented Architecture (SOA) applications, in domains such as financial services, manufacturing, gaming and military/aerospace, have real-time performance requirements. We present real-life industry use cases from these domains as motivation for the potential benefit in developing real-time complex event processing applications. In support of a feasibility argument for the proposed approach we present some preliminary experimental results obtained on a partially implemented tool.

Active database systems are systems that detect events and trigger actions as a result of this detection. Active capabilities are provided by a set of rules, such that each rule consists of three components (event, condition, and action). A major performance issue in active databases is the issue of relationships among rule components. Current implementations of triggers do not allow flexibility in the selection of transaction policies (partition of rules to transactions); the intertransaction timing policies of rules' components, the intratransaction policies of commit and abort dependencies, and synchronization issues. While these decisions have a substantial impact on the application performance, they are not provided as design primitives; one of the reasons for that is that it is very difficult to manually tune these decisions. In some research prototypes of active databases, these relationships are encapsulated into a set of coupling modes. Each coupling mode represents a combination of decisions about the partition of rule components to transactions, the relative timing within a transaction, and the interrelationships among these transactions. We describe a self-tuning model that operates on a general active database. The optimization model strives to minimize a programmable goal function that reflects the system designer's preferences and the system behavior and the applications' semantics through constraint definitions. The tuning model strives to optimize the mutual relationships among the system rules' components.

Events are at the core of reactive and proactive applications, which have become popular in many domains.

The research in materialization of derived data elements has dealt so far with the if issue of whether to physically store derived data elements. In the active database area, there has been some research on the how issue. We deal with the when issue, devising an optimization model to determine the optimal materialization strategy. The decision problem confronted by the optimization model deals with devising the materialization strategy that consists of a set of interdependent decisions about each derived data element. Each decision relates to two issues: Should the value of a derived data element be persistent? What is the required level of consistency of a derived value with respect to its derivers? For each derived data element, the decision is based on both its local properties and its interdependencies with other derived values. The optimization model is based on a heuristic algorithm that finds a local optimum in O(N/sup 2/) and a monitor that obtains feedback about the actual database performance. This optimization model is general and is not specific to any data model. Our experimental results show that a predictor for the optimal solution cannot be obtained in any intuitive or analytic way, due to the complexity of the involved considerations; thus, there is no obvious way to achieve these results without using the optimization model. This fact is a strong motivation for applying such an optimization model.