Business Process Management (BPM) and Process Aware Information Systems (PAIS) are becoming more and more integrated in todays business environments. The formulation of business strategies leads to business requirements for business processes. These business processes are designed, modeled and documented using formal, commercial or ad-hoc modeling languages. Furthermore, some modeled business processes get translated to executable entities, enabling the automatic enactment and follow-up of a business process. Especially in combination with Service Oriented Architectures (SOA) an enactment environment is created where processes can be deployed and executed automatically. From a managerial and technical point of view, the interpretation, control and enactment of a business process's control flow happens very often at one point in the organizational and IT structure. This creates an inflexible environment, where control over and visibility of cross-departmental processes cannot be distributed across these organizational entities. Moreover, a centralized approach creates a performance bottleneck and single point of failure in process execution. As the number of process instantiation requests increases, time to successfully handle one process instance also increases. Process model fragmentation and distribution is a technique to overcome the aforementioned issues. Process model fragmentation is the process of splitting a process model that was modeled as a whole into logically different, smaller model fragments with the intention to distribute the fragments over different execution and controlling partners. This dissertation studies the challenging aspects of process fragmentation and proposes a non-intrusive, automatic approach to fragment and distribute the process flow over different organizational entities, hereby also increasing performance and removing any single point of failure. Special attention is given to flexibility: each distributed process part becomes autonomous, creating the ability to easily add new process fragments, change the deployment structure and add monitoring and management tools without any additional effort.The first part of this dissertation discusses the fragmented enactment environment, the architecture, transformations and semantics needed to deploy and run a fragmented process model. The second part focuses on runtime adaptability of the process model in the fragmented and distributed environment. Similar to centralized process enactment, the distributed environment should support the ability to respond effectively to process changes. In the second part the difficulties, advantages and issues of process model change support are discussed. Moreover a system is proposed which tackles the identified issues and allows the propagation and coordination of process changes in the distributed process execution architecture. Two change types are discussed: top-down and bottom up. The top-down change is issued by a central coordinator, given the originally designed process model. The bottom-up change is issued by alocal fragment coordinator, given only the information available in his own process model fragment. Both thesis parts are supported by a formal and experimental verification and a proof-of-concept implementation, showing the feasibility of the runtime and adaptability potential of the fragmented process execution environment.
展开▼
