Cloud service brokerage leads to creation of ecosystems of highly distributed, task-oriented, modular, and collaborative cloud services managed by a broker. A broker is striving to create optimized cloud service consumption lifecycle in terms of cost, flexibility and business agility. In order to effectively manage the complexity inherent in such ecosystems, enterprises are anticipated to crucially depend upon cloud service brokerage (CSB) mechanisms. This work focuses on the management of hosting platforms participating in the ecosystem of a cloud service brokerage platform. The hosting platforms are as any other actor of a cloud service brokerage ecosystem evolving over time. The hosting platforms may join or leave the ecosystem, add or remove hosting services to the ecosystem or change characteristics of the available hosting services. The broker is thereby confronted with the issue of keeping its business policy offered to the service providers and service consumers up to date concerning the hosting alternatives in the ecosystem. We present a strategy for derivation of business policies from service descriptions of hosting services. The strategy is showcased in Linked USDL – our chosen technical specification for enabling platform-agnostic data exchanges.

In future Cloud ecosystems, brokers will mediate between service providers and consumers, playing an increased role in quality assurance, checking services for functional compliance to agreed standards, among other aspects. To date, most Software-as-a-Service (SaaS) testing has been performed manually, requiring duplicated effort at the development, certification and deployment stages of the service lifecycle. This paper presents a strategy for achieving automated testing for certification and re-certification of SaaS applications, based on the adoption of simple state-based and functional specifications. High-level test suites are generated from specifications, by algorithms that provide the necessary and sufficient coverage. The high-level tests must be grounded for each implementation technology, whether SOAP, REST or rich-client. Two examples of grounding are presented, one into SOAP for a traditional web service and the other into Selenium for a SAP HANA rich-client application. The results demonstrate good test coverage. Further work is required to fully automate the grounding.

With the rise of cloud computing, enterprises increasingly rely for their daily operations on heterogeneous externally-sourced cloud services that span different levels of capability. Their IT environment is thus progressively transformed into an ecosystem of intertwined infrastructure, platform, and application services. To effectively manage the ensuing complexity, enterprises are anticipated to increasingly rely on cloud service brokerage (CSB). This work presents a conceptual architecture for a framework which provides solutions with respect to the quality assurance and optimisation dimensions of CSB in the context of virtual enterprises. The framework revolves around three general themes, namely governance and quality control, failure prevention and recovery, and optimisation.

As the number of cloud service providers grows and the requirements of cloud service consumers become more complex, the latter will come to depend more and more on the intermediation services of cloud service brokers. Continuous quality assurance and optimisation of services is becoming a mission-critical objective that many consumers will find difficult to address without help from cloud service intermediaries. The Broker@Cloud project envisages a software framework that will make it easier for cloud service intermediaries to address this need, and this paper provides an analysis of key requirements for this framework. We discuss the methodology that we followed to capture these requirements, which involved defining a conceptual service lifecycle model, carrying out a series of Design Thinking workshops, and formalising requirements based on an agile requirements information model. Then, we present the key requirements identified through this process in the form of summarised results.

With the pervasion of cloud computing, enterprises increasingly rely on ecosystems of distributed, task-oriented, modular, and collaborative cloud services. In order to effectively manage the complexity inherent in such ecosystems, enterprises are anticipated to depend upon brokerage mechanisms for performing policy-based governance and for recommending optimal services to consumers. Such mechanisms crucially depend upon the existence of a uniform, platform-independent representation of services, consumer preferences, and policies concerning service delivery. In this paper we propose an ontology-based approach to such a representation.

In process engineering, processes can be refined from simple ones to more and more complex ones with decomposition and restructuring of activities. The validation of these refinements and the explanation of invalid refinements are non-trivial tasks. This paper formally defines process refinement validation based on the execution set semantics and presents a suite of refinement reduction techniques and an ontological representation of process refinement to enable reasoning for the validation and explanation of process refinement. Results show that it significantly improves efficiency, quality and productivity of process engineering.

Cloud service intermediation is becoming increasingly recognized as a key component of the cloud computing value chain. Existing cloud service intermediaries already offer capabilities such as integration, customization or aggregation brokerage, but in the future, enterprises will require much more sophisticated capabilities going far beyond what is available today. The types of brokerage capabilities foreseen to be most valuable for service consumers, and at the same time most demanding for future enterprise service brokers to implement, are those addressing continuous quality assurance and optimization of cloud services. In this paper, we present a research roadmap for bringing continuous quality assurance and optimization capabilities closer to the reach of enterprise cloud service brokers. We present a scenario for motivating the need for such capabilities, as well as a discussion of the dominant views on cloud service brokerage.

Business today is increasingly becoming servicedriven. A service represents either a part of or a complete business process, which in turn depicts the life cycle of a Business Object (BO). A number of actions constrained by a set of business policies cause the BO to transit from an initial state to a final state during its life cycle. The constraints would vary for different customized business processes. To engineer reliable service based systems it is crucial to verify the reachability of the desired final states of the BO model against a given set of constraints. We propose the formal specification of a constraint-driven BO model, and thereby, present a verification tool to validate the life cycle of the BO model given a set of constraints. The core object and its constraint-driven life cycle are modeled separately for flexibility and customizability. We also define a language specification to represent our BO model and the business goals. The specification is automatically interpreted by our verification tool, which outputs a graphical view of the life cycle and a textual report of the validation results.

A prominent research focus, especially in the context of EU public funding, has been the systematic use of the Internet for new ways of value creation in the services sector. This idea of service networks in the Internet, frequently dubbed the Internet of Services orWeb service ecosystems, wants to make services tradable in digital media. In order to enable communication and trade between providers and consumers of services, the Internet of Services requires a standard that creates a “commercial envelope” around a service. This is where the Unified Service Description Language (USDL) comes into play as a normative and balanced unification of service information. The unified description established by USDL is machineprocessable, considers technical and business aspects of a service as well as functional and non-functional attributes.

This chapter provides a formal specification of non-atomic, relaxed action refinement suited for component-based business process engineering. Engineering a business process involves multiple process models created by different people on different levels of abstractions. Keeping the models consistent during the engineering procedure-refinement validation-is one objective of this chapter. In component-based software engineering, the lowest abstraction of a business process is mapped on existing components that have a description of their behaviors. Checking the consistency of process and component behavior-grounding validation-is the second objective. Both refinement and grounding validation increase the robustness of business process implementations and the productivity of process engineers. Technically, the specification given in this chapter is in terms of deadlock analysis in safe Petri nets. The evaluation of this straight-forward implementation underlines the exponential complexity of deadlock analysis in safe Petri nets. For use cases with more than 30 activities per process or heavy parallelism, optimized implementations are needed.