Architectural support for dynamic reconfiguration of large scale distributed applications

In a distributed environment, it is inevitable that long-running applications will require support for dynamic reconfiguration because, for example, machines may fail, services may be moved or withdrawn and user requirements may change. In such an environment, it is essential that the structure of running applications can be modified to reflect such changes. A complication is that such long-running applications are frequently composed our of existing applications. The resulting application can be very complex in structure, containing many temporal dependencies between constituent applications. This paper describes an approach that supports the dynamic reconfiguration of large-scale distributed applications. An application composition and execution environment has been designed and implemented as a transactional workflow system that enables sets of inter-related tasks (applications) to be carried out and supervised in a dependable manner. A task model that is expressive enough to represent temporal dependencies between constituent tasks has been developed. The workflow system maintains this structure and makes it available through transactional operations for performing changes to it. Use of transactions ensure that changes can be carried out atomically with respect to running applications. The workflow system is general purpose and open: it has been designed and implemented as a set of CORBA services to run on top of a given ORB