Robin Fischer

Scalable Business Process Execution in the Cloud

Business processes orchestrate service requests in a structured fashion. Process knowledge, however, has rarely been used to predict and decide about cloud infrastructure resource usage. In this paper, we present an approach for BPM-aware cloud computing that builds on process knowledge to improve the timeliness and quality of resource scaling decisions. We introduce an IaaS resource controller based on fuzzy theory that monitors process execution and that is used to predict and control resource requirements for subsequent process tasks. In a laboratory experiment, we evaluate the controller design against a commercially available state-of-the-art auto scaler. Based on the results, we discuss improvements and limitations, and suggest directions for further research.

Perspectives for Web Service Intermediaries: How Influence on Quality Makes the Difference

In the service-oriented computing paradigm and the Web service architecture, the broker role is a key facilitator to leverage technical capabilities of loose coupling to achieve organizational capabilities of dynamic customer-provider-relationships. In practice, this role has quickly evolved into a variety of intermediary concepts that refine and extend the basic functionality of service brokerage with respect to various forms of added value like platform or market mechanisms. While this has initially led to a rich variety of Web service intermediaries, many of these are now going through a phase of stagnation or even decline in customer acceptance. In this paper we present a comparative study on insufficient service quality that is arguably one of the key reasons for this phenomenon. In search of a differentiation with respect to quality monitoring and management patterns, we categorize intermediaries into Infomediaries, e-Hubs, e-Markets and Integrators. A mapping of quality factors and control mechanisms to these categories depicts their respective strengths and weaknesses. The results show that Integrators have the highest overall performance, followed by e-Markets, e-Hubs and lastly Infomediaries. A comparative market survey confirms the conceptual findings.

Compliance-Preserving Cloud Storage Federation Based on Data-Driven Usage Control

Cloud storage federation improves service availability and reduces vendor lock-in risks of single-provider cloud storage solutions. Federation therefore distributes and replicates data among different cloud storage providers. Missing controls on data location and distribution however introduce security and compliance issues. This paper proposes a novel approach of using data-driven usage control to preserve compliance constraints in cloud storage federation. Based on common compliance regulations and laws we provide a brief categorization of compliance problems into spatial, temporal, and qualitative requirements. In addition, we show how usage control policies can be employed to constrain federation according to these categories. To demonstrate the feasibility of our approach we evaluate security and performance of our prototypical implementation.

Towards an Automated Gap Analysis for E-Service Portfolios

Intermediaries for e-services continuously gain momentum, powered by a materializing Internet of Services. However, quality of service still exhibits considerable shortcomings, as no structured process to enhance consumer satisfaction is available yet. To improve the match of delivered e-service quality and expected service quality on the consumer side, we develop a portfolio optimization process that integrates both, the consumers as well as the intermediarys perspective. First, we introduce a toolkit for an e-service oriented gap analysis. Thereupon, we identify monitoring points to measure service quality gaps automatically. A subsequent aggregation of measured data into customized feedback information allows for applying the toolkit to continuously optimize e-service portfolios. Instantiated in the AGORA e-service market, we conclude with a report on our recent implementation results.

Consistency Benchmarking: Evaluating the Consistency Behavior of Middleware Services in the Cloud

Cloud service providers such as Amazon Web Services offer a set of next-generation storage and messaging middleware services that can be utilized on-demand over the Internet. Outsourcing software into the cloud, however, confronts application developers with the challenge of understanding the behavior of distributed systems, which are out of their control. This work proposes an approach to benchmark the consistency behavior of services by example of Amazon Simple Queue Service (SQS), a hosted, Web-scale, distributed message queue that is exposed as a Web service. The data of our consistency benchmarking tests are evaluated with the metric harvest as described by Fox and Brewer (1999). Our tests with SQS indicate that the client-service interaction intensity has an influence on harvest.

DYNO: A notation to leverage dynamic network effects in PaaS ecosystems

Platform-as-a-service offerings continuously gain importance as two-sided markets, offering Software-as-a-Service (SaaS) to the respective customers. Market success is achieved by platforms, which excel in shaping ecosystems of users and autonomous SaaS suppliers around their basic value proposition — and in controlling quality of service in function of customer requirements. In this paper, we suggest a notation for dynamic networks ‘DYNO’, designed to help platform providers in creating PaaS, optimized on their specific requirements. In addition, DYNO gives support in properly allocating control mechanisms to guarantee high quality of service. In a use-case we describe how business analysts and service engineers may use DYNO-models to conceive PaaS ecosystems.

Composite Solutions for Consumer-Driven Supply Chains

Global economy migrates from vertically integrated enterprises towards specialized enterprises interoperating to create end-to-end value to customers. Value nets, emerging around the value proposition of a platform offer, have become the centerpiece of collaborative value creation, altering the traditional supply chain approach, while keeping pace with changing customer demands. This modular approach allows for sophisticated, cost efficient composite solutions provided by dynamic ‘consumerdriven supply chains’. We characterize these supply chains ‘consumer-driven’ as the services complementing the platform offer are built based on loosely coupled supply chains of services, provided by an ‘open pool of autonomous service providers’, which we will refer to as ‘service-enabling ecosystem’. In this paper, we describe the challenges platform operators face in the light of consumer-driven supply chains. In particular, we focus on how to orchestrate a complex self-organizing service-enabling ecosystem, on how to steer a continuous service supply and explicitly, on how to ensure sustainable service quality optimization. Based on a conceptual model of feedback control, which we introduce in pursuit of the methods of control engineering, we propose a managerial control process for service-enabling ecosystems as well as a corresponding typology of control mechanisms for improved consumer-driven service supply on top of a platform.

A Life-Cycle Model for Software Service Engineering

Applying existing life-cycle models to software service engineering, we identify shortcomings: they do not focus on engineering activities, lack a clear underlying definition of software services, do not address both service roles of consumer and provider, and assume longevity and sequence of activities. We present a life-cycle model for software service engineering to tackle these shortcomings. We illustrate the model’s prescriptive use by applying it to assess software service variability and software service standardization.

The dynamic network notation: harnessing network effects in PaaS-ecosystems

Web applications complement the Platform-as-a-Service (PaaS) value by satisfying widespread and rapidly changing consumer requirements within limited time and budget. Successful PaaS providers excel in governing their market performance by leveraging complex network effects, which implicitly control PaaS-ecosystems. There is currently no methodically sound and easy to use tool available to business analysts and software engineers of PaaS-offerings that addresses challenges and opportunities in launching and governing such highly dynamic networks. In this paper, we capture network behavior through elements of complex system and control theory. Our dynamic network notation (DYNO) builds upon these theories. In more detail, DYNO models PaaS offerings with a focus on identifying and shaping network effects towards a sufficient user-base and an optimized portfolio of Web applications, all while maintaining a high quality of service.

Business activity management for service networks in cloud environments

Companies struggle to find ways to manage intra- and interorganizational service networks communicating in a distributed fashion across the globe. We review the state-of-the-art of managing choreographed service networks and put it in relation to process analytics and complex event processing (CEP) against the background of Cloud computing. We present an initial architecture for Event-driven Business Activity Management of service networks which also takes into consideration levels of virtualization. The architecture can serve as a blueprint for flexible business activity monitoring applications as well as closed loop service choreography control solutions. We illustrate the interaction of Cloud infrastructure, services networks, and CEP systems with a number of use cases. In addition, we discuss future research directions based on our experiences from early prototypes.