Run-time configuration management specification for distributed systems

Two methods are introduced; one that identifies key run-time configuration management requirements in complex distributed systems, and another that specifies these key requirements, validates their adequate representation in a specification, and verifies specification internal consistency. The methods are created using experiential observations of the development of distributed systems. They are generalized to work with distributed systems in other problem domains. Techniques used in industry are extended and combined with techniques proposed by academia to create specifications that can be validated for adequate representation of key run-time configuration management requirements as well a for internal consistency. New methods and notations combine industrial and academic methods to adequately capture flexibility and extensibility requirements in the specification of software run-time configuration management for complex distributed systems. The methods focus on software items involved in run-time configuration management.; Specification methods used in industry are not expressive enough to precisely specify key software run-time configuration management requirements of loosely coupled distributed systems. This research analyzes a system that has required large amounts of engineering resources to maintain and enhance, in part as a consequence of not meeting its original run-time configuration management flexibility and extensibility requirements. Methods are developed to adequately capture key software run-time configuration managements of this system, to validate their existence in a specification, and to verify specification internal consistency. These methods are then generalized and demonstrated on a distributed system in another problem domain. The methods are developed to be acceptable for use in an industrial software development setting. Methods are scalable to large systems since they were developed using a large system as the case study. They do not require extraordinary skills to use and are flexible to fit into the existing development paradigm of an organization.