Dynamic WSDL for Supporting Autonomic Computing

Abstract

An autonomic computing system is organized into building blocks that can be composed together to form a self-managing system. Architecturally, this matches service-based computing systems, which are the outcomes of the most recent effort to provide interoperability and usability through the use of services. Autonomic computing is an attractive information technology for managers and clients. Rather than overwhelm administrators and programmers with hundreds or even thousands of machines within a distributed computing system, a cloud, autonomic computing systems manage themselves thus freeing some of the underlying tasks from administrators and programmers. When the new component is added, the system learns of it and makes use of the component by itself. When a component fails, attempts are made to automatically recover the component before human assistance is requested. But to support such complexity, an environment rich in current contextual information is needed. Currently, this information has to be extracted via notifications, which themselves require the discovery of a notification management services. For service-based distributed systems, the knowledge of service state and its attributes is of the most crucial importance for the service provision. We propose the application of the innovative Resources Via Web Instances framework, which allows the state of services to be exposed via their interfaces. The inclusion of state in the interface directly crates an information rich context for autonomic distributed systems without additional complexity of notification mechanics. Furthermore, our innovative framework allows the attributes of a web service to be shown, thus allowing the autonomic system to better cater for the installation and use of new components. With the features of a new component shown as attributes, the autonomic distributed system can take into consideration the nature of the new component.