Arun Anandasivam

Cloud Computing – A Classification, Business Models, and Research Directions

Lately, a new computing paradigm has emerged: “Cloud Computing”. It seems to be promoted as heavily as the “Grid” was a few years ago, causing broad discussions on the differences between Grid and Cloud Computing. The first contribution of this paper is thus a detailed discussion about the different characteristics of Grid Computing and Cloud Computing. This technical classification allows for a well-founded discussion of the business opportunities of the Cloud Computing paradigm. To this end, this paper first presents a business model framework for Clouds. It subsequently reviews and classifies current Cloud offerings in the light of this framework. Finally, this paper discusses challenges that have to be mastered in order to make the Cloud vision come true and points to promising areas for future research.

Environment-conscious scheduling of HPC applications on distributed Cloud-oriented data centers

The use of High Performance Computing (HPC) in commercial and consumer IT applications is becoming popular. HPC users need the ability to gain rapid and scalable access to high-end computing capabilities. Cloud computing promises to deliver such a computing infrastructure using data centers so that HPC users can access applications and data from a Cloud anywhere in the world on demand and pay based on what they use. However, the growing demand drastically increases the energy consumption of data centers, which has become a critical issue. High energy consumption not only translates to high energy cost which will reduce the profit margin of Cloud providers, but also high carbon emissions which are not environmentally sustainable. Hence, there is an urgent need for energy-efficient solutions that can address the high increase in the energy consumption from the perspective of not only the Cloud provider, but also from the environment. To address this issue, we propose near-optimal scheduling policies that exploit heterogeneity across multiple data centers for a Cloud provider. We consider a number of energy efficiency factors (such as energy cost, carbon emission rate, workload, and CPU power efficiency) which change across different data centers depending on their location, architectural design, and management system. Our carbon/energy based scheduling policies are able to achieve on average up to 25% of energy savings in comparison to profit based scheduling policies leading to higher profit and less carbon emissions.

Business Models in the Service World

The authors provide a criteria catalogue to characterize cloud computing and their own Cloud Business Ontology Model to classify current product offerings and pricing models.

SORMA - building an open grid market for grid resource allocation

The demand for computing and storage resources in a Grid network increases in both academic and industrial application domains. Participants in a network (i.e. companies or research institutes) try to selfishly maximize their individual benefit from participating in the Grid. Setting the right incentives for suppliers and requesters for an efficient usage of the limited Grid resources will motivate the participants to cooperate and provide their idle resources. In this paper we present an economic approach for efficient resource allocation. A market mechanism called Decentralized Local Greedy Mechanism [2] satisfies desirable economic properties and thus is deemed promising to enable an efficient allocation of Grid resources.

SORMA --- Business Cases for an Open Grid Market: Concept and Implementation

Economic mechanisms enhance technological solutions by setting the right incentives to reveal information about demand and supply accurately. Market or pricing mechanisms are ones that foster information exchange and can therefore attain efficient allocation. By assigning a value (also called utility) to their service requests, users can reveal their relative urgency or costs to the service. The implementation of theoretical sound models induce further complex challenges. The EU-funded project SORMA analyzes these challenges and provides a prototype as a proof-of-concept. In this paper the approach within the SORMA-project is described on both conceptual and technical level.

Software als Institution und ihre Gestaltbarkeit

ZusammenfassungSoftware regelt immer mehr zwischenmenschliche Interaktionen. Üblicherweise werden die Funktionsmechanismen, Wirkungen und Gestaltungsoptionen von Regeln in der Institutionenforschung behandelt. In diesem Artikel soll beleuchtet werden, inwieweit sich Ansätze der Institutionenforschung auf Software anwenden lassen und was sich aus dieser Forschungsperspektive zu den Regelungswirkungen und Gestaltungsoptionen von Software ableiten lässt.

Managing Revenue in Grids

The distributed usage of computing resources over a large-scale network allows users to receive and offer resources on demand. The on demand paradigm leads to dynamic and unpredictable usage of resources, since every user in the network will try to maximize his utility by selfish behavior. The customers behavior can be actuated by pricing policies to lower demand at peak time. Revenue Management as a relatively new economic paradigm provides various tools to optimally allocate capacity and increase revenue. We provide a framework how the matured concepts of Revenue Management can be deployed to Grid Computing. We analyze whether the Grid Computing domain has notable differences from the airline industry or other common areas for Revenue Management like restaurant, hotel or car rental industries. Hence, we outline tools and methods known from Revenue Management and how they can be applied to Grid Computing.

A Heuristic Approach for Capacity Control in Clouds

Cloud resource providers in a market face dynamic and unpredictable consumer behavior. The way, how prices are set in a dynamic environment, can influence the demand behavior of price sensitive customers. A Cloud resource provider has to decide on how to allocate his scarce resources in order to maximize his profit. The application of bid price control for evaluating incoming service requests is a common approach for capacity control in network revenue management. In this paper we introduce a customized version of the concept of selfadjusting bid prices and apply it to the area of Cloud Computing. Furthermore, we perform a simulation in order to test the efficiency of the proposed model.

Enabling Cloud Service Reservation with Derivatives and Yield Management

Since the cloud computing advent a few years ago this paradigm has proved itself to be extremely successive in the industrial sector. IT and business managers agree that this form of highly-dynamic provision of IT services will play a major role in the IT world during the next decades. Several business models have already been developed and implemented, including also dynamic pricing schemes. Given that, we analyze in this work two distinct reservation system approaches from related areas and how they can be applied to the cloud computing scenario. We state the different requirements and set up a realistic implementable model in order to enable cloud vendors to improve their revenues.

Managing the Quality of Modular Services -- A Process-Oriented Aggregation of Expected Service Levels Based on Probability Distributions

Service oriented architectures provide modular service components which can be combined to complex services. Besides the problem of finding optimal combinations of service components the quality of the overall service (the complex service) is important. It is often assumed in literature that the quality of the individual service components is expressed as one single value (e.g. the average). In fact, we observe different probabilities for certain quality levels which are summarized in a probability distribution of the service level indicators. Given the distribution of the quality levels of the individual service components we present an approach to aggregate these distribution functions. As such, we are able to derive one distribution function for the overall complex service. The distribution function of the complex service carries much more information than a single value. This fact is essential for analyzing, managing and controlling the overall process in order to detect weaknesses and to approve overall quality.