Accountability in cloud services
thesisposted on 2022-03-28, 00:59 authored by Jun Zou
Service computing has become the main theme in IT while Web services and cloud services are widely adopted by the industry. An important characteristic of service computing is that it turns IT capability into a service, and facilitates automatic service delivery and consumption through the ubiquitous Internet. As such, it closely aligns IT with business, which brings up the important issue of accountability that is largely overlooked in traditional IT. In a business context, accountability encompasses transparency, responsibility, responsiveness and willingness for assuming liability. Applying that concept in a service computing context, it should mean a clear disclosure of service obligations; faithfully honoring disclosed obligations, or otherwise assuming liability for the unsatisfactory performance of the obligations. A comprehensive study of the accountability literature reveals that the vast majority of researchers in the IT community share a quite different view on accountability than their counterparts in the business community. Most researchers in IT tend to take some aspects of quality of service (QoS) or architectural concerns, such as security, provenance and auditability, as accountability, missing the crucial components of disclosure, obligation fulfilment monitoring and liability assignment that concern business more frequently. This discrepancy in understanding can cause a significant gap between business’s expectations of accountability and the actual accountability capability of an IT system. While this gap may not manifest as a major concern in the traditional IT environment, it will exert a serious negative impact on the development of service computing, since a major theme of service computing is the close alignment of IT and business. Three main objectives are set in this thesis. The first is to raise awareness of the accountability gap from both the conceptual perspective and the actual architectural implementation perspective. The second is to clarify the confusion on the topic of accountability in the service computing industry and, more importantly, to lay down a foundation to strengthen accountability in service computing. The third is to describe approaches for building an advanced service accountability mechanism, aimed at automating the accountability processes in a truly service-oriented environment. Accordingly, the thesis is structured in a way that progressively meets the above objectives. The Introduction chapter focuses on achieving the first objective, and sets the theme for the rest of the thesis. The second objective, is achieved through two steps. The first step is to clarify the confusion around the topic of accountability through a thorough analysis of the existing accountability literature. The second step involves a series of tasks for laying down the service accountability foundation by proposing a service accountability framework, addressing the accountability weaknesses of the current Service-Oriented Architecture. The third objective, is achieved through three different approaches. The first approach is to represent a service contract using semantic web technology. The second approach is to represent a service contract using dynamic logic and process algebra techniques. The third approach provides a decentralised service contract management scheme based on the blockchain technology, taking advantage of the irreversibilityand tamper-proof features of the blockchain, and presents a scheme for service contract disclosure and obligation tracking. We hope that through achieving these three objectives, a solid foundation can be laid for strengthening accountability in service-oriented environments, which addresses business’s concerns on service accountability and boosts business’s confidence in fully embracing cloud services.