Implementing an Active Functionality Service on Different Platforms
Jose Antollini · TUbilio (Technical University of Darmstadt) · 2005
Business requirements are changing faster than applications can be created and/or adapted. Most of these requirements are in the form of or are related to business rules. Business rules are precise statements that describe, constrain and control the structure, operations and strategy of a business. They offer a way of encapsulating business semantics and making them explicit. Traditionally, business rules have been scattered, hard-coded and replicated by different applications. As a result, it has been difficult to adapt applications to new requirements quickly. In recent years, there has been a trend to extend database technology with powerful rule-processing capabilities leading to the emergence of so-called active databases (aDBMS for short). They rely on Event-ConditionAction rules (ECA-rules) to provide automatic execution of predefined operations in response to the occurrence of certain events. This powerful technology is particularly convenient for enforcing business rules. Modern large-scale applications, such as e-commerce, Internet or Intranet applications, enterprise application integration (EAI), and emerging pervasive systems, can benefit from an active mechanism. But they also impose new requirements. These applications demand support for: information integration, interaction with different systems or services, and collaboration among applications. In this context and from the active database perspective, events and data come from diverse sources, and the execution of actions and evaluation of conditions may be performed on different systems. Conventional active mechanisms have been designed for centralized systems and they are monolithic making it difficult to extend and adapt. Consequently, active mechanisms must be adapted to meet the requirements imposed by this generation of large-scale heterogeneous applications. That is why, for these new business strategies an appropriate reactive functionality must be devised. Because of they depend on the reaction to happenings and efficient event dissemination related to their business rules. The current trend in the application space is moving towards systems of loosely-coupled, dynamically bound components. In such a context, it seems reasonable offering an autonomous service that can be combined in many different ways and used in a variety of environments. Moreover, in a distributed and heterogeneous environment active functionality mechanisms are fed with events coming from heterogeneous sources. These events encapsulate data, which can only be properly interpreted when sufficient context information about its intended meaning is known. In general, this information is left implicit and as a consequence, it is lost when data/events are exchanged across institutional or system boundaries. For this reason, to exchange and process events from independent sources in a semantically meaningful way, explicit information about their semantics in the form of additional meta-data is required. For this purpose, these new ways of interaction that are present in emerging applications need the adequate treatment by means of appropriate interaction mechanisms. They should be based