Florian Niedermeier

Performance tradeoffs of energy-aware virtual machine consolidation

Increasing power consumption of IT infrastructures and growing electricity prices have led to the development of several energy-saving techniques in the last couple of years. Virtualization and consolidation of services is one of the key technologies in data centers to reduce overprovisioning and therefore increase energy savings. This paper shows that the energy-optimal allocation of virtualized services in a heterogeneous server infrastructure is NP-hard and can be modeled as a variant of the multidimensional vector packing problem. Furthermore, it proposes a model to predict the performance degradation of a service when it is consolidated with other services. The model allows considering the tradeoff between power consumption and service performance during service allocation. Finally, the paper presents two heuristics that approximate the energy-optimal and performance-aware resource allocation problem and shows that the allocations determined by the proposed heuristics are more energy-efficient than the widely applied maximum-density consolidation.

A methodology to predict the power consumption of servers in data centres

Until recently, there have been relatively few studies exploring the power consumption of ICT resources in data centres. In this paper, we propose a methodology to capture the behaviour of most relevant energy-related ICT resources in data centres and present a generic model for them. This is achieved by decomposing the design process into four modelling phases. Furthermore, unlike the state-of-the-art approaches, we provide detailed power consumption models at server and storage levels. We evaluate our model for different types of servers and show that it suffers from an error rate of 2% in the best case, and less than 10% in the worst case.

Renewable energy-aware data centre operations for smart cities the DC4Cities approach

Data centres are important players in smart cities both as IT service providers and as energy consumers. Integrating intermittent renewable energy sources into the local power grid is one challenge in future smart cities aiming at an IT based low carbon economy. The project DC4Cities takes up this challenge by offering a both technical and business related solution for optimizing the share of local renewable power sources when operating data centres in smart cities. To this end, power management options between the data centre and the smart city together with internal adaptation strategies for data centres are introduced. Finally, an implementation of the suggested approach is presented and evaluated in a simulation.

Quality Assessment of the MPEG-4 Scalable Video CODEC

In this paper, the performance of the emerging MPEG-4 SVC CODEC is evaluated. In the first part, a brief introduction on the subject of quality assessment and the development of the MPEG-4 SVC CODEC is given. After that, the used test methodologies are described in detail, followed by an explanation of the actual test scenarios. The main part of this work concentrates on the performance analysis of the MPEG-4 SVC CODEC - both objective and subjective. Please note that this document is only a shortened version of the assessment. Further experimental results can be found in the extended version available at the Computing Research Repository (CoRR).

Greensdas leveraging power adaption collaboration between energy provider and data centres

Data centres, due to their significant energy use and highly automated IT infrastructure, are excellent candidates to participate in demand response programs. However, the major inconvenience of todays electricity tariffs between energy provider and its customers is their lack of flexibility, which renders demand response programs difficult to realise. In this paper, we propose a new supply demand agreement in order to foster power adaption (i.e. increase/decrease) collaboration between energy provider and data centres. To this end, we introduce contractual terms and based on those we propose reward and penalty schemes. Furthermore, we provide a signalling scheme which defines the communication requirements necessary to enable power adaption collaboration. Finally, we present a scheduling policy which helps the energy provider to request data centres in a fair manner for power adaption.

Introducing Flexibility into Data Centers for Smart Cities

In Europe, more and more cities are aiming to become part of the “smart cities” vision. Smart Cities are based on a plethora of sensor data which need to be processed in data centres. Therefore data centres play an important role in making cities smart. However, at the same time they are huge consumers of electrical energy and thus counteract smart cities’ goals of an IT based low carbon economy. The project DC4Cities takes up the challenge of turning data centres into flexible energy consumers that to a high degree run on renewable energy sources. It offers a technical solution for optimizing the share of renewables in data centre energy consumption and supports this by novel contracts and business models. This paper introduces power management options between data centres and a smart city which are backed by adaptation strategies within the data centre. It also presents a set of contracts that complement the technical solution and a trial evaluation of the approach.

Greener Bits: Formal Analysis of Demand Response

Demand response is a promising approach to deal with the emerging power generation fluctuations introduced by the increasing amount of renewable energy sources fed into the grid. Consumers need to be able to adapt their energy consumption with respect to the given demand pattern and at the same time ensure that their adaptation (i.e., response) does not interfere with their various operational objectives. Finding, evaluating and verifying adaptation strategies which aim to be optimal w.r.t. multiple criteria is a challenging task and is currently mainly addressed by hand, heuristics or guided simulation. In this paper we carry out a case study of a demand response system with an energy adaptive data center on the consumer side for which we propose a formal model and perform a quantitative system analysis using probabilistic model checking. Our first contribution is a fine-grained formal model and the identification of significant properties and quantitative measures (e.g., expected energy consumption, average workload or total penalties for violating adaptation contracts) that are relevant for the data center as an adaptive consumer. The formal model can serve as a starting point for the application of different formal analysis methods. The second contribution is an evaluation of our approach using the prominent model checker PRISM. We report on the experimental results computing various functional properties and quantitative measures that yield important insights into the viability of given adaptation strategies and how to find close-to-optimal strategies.

Energy supply aware power planning for flexible loads

Increasing the use of renewable energy is considered a viable way of reducing carbon intensive power generation. However, a power grid running on high amounts of renewable energy has to deal with the limited controllability and higher volatility of power sources like wind or solar. In this work, we propose to use demand side management to deal with varying amounts of renewable power feed-in via the use of power plans, i.e. instructions passed to large energy consumers that specify how they should try to spread out their energy use over a day. We argue that a separation of power planning and implementation of technical measures to schedule loads to follow the plan would alleviate some of the problems faced by an integrated planning-scheduling approach, as these processes are governed by different entities who may be unwilling to disclose all required information to each other. As a proof-of-concept, we propose and analyze a quadratic programming approach to maximizing the fraction of renewable energy being used while not overburdening the consumer with a power plan that is difficult to follow.

Demand-Side Flexibility and Supply-Side Management: The Use Case of Data Centers and Energy Utilities

Lately power grids have been subject to one of their major evolutions since their design and conception. The traditional structure of electricity being generated by a small number of huge and centralized power plants is being defied by the increasing penetration of renewable energy sources. The major drawback of such sources is their intermittent behavior rendering power generation planning even more cumbersome. This problem can be alleviated through the implementation of intelligent energy management systems (EMS) whose main objective is to exploit demand-side flexibilities for the purpose of better supply-side management and planning. Data centers, on one hand due to their significant power (in the order of up to 200 MW) as well as energy demand, and on the other hand thanks to their highly automated ICT infrastructure providing flexibilities without human interventions, have been shown to be excellent candidates for participation to such EMS. To this end, in this chapter we study such energy management systems by considering the use case of data centers both from local as well as coordinated management perspectives. For each considered perspective we describe thoroughly the concept as well as give its corresponding architectural building blocks. Furthermore, we specify the mechanisms and strategies that can be used for the case of data centers in exploiting demand-side flexibilities.

Mapping of Self-organization Properties and Non-functional Requirements in Smart Grids

Future electrical power networks will be composed of large collections of autonomous components. Self-organization is an organizational concept that promises robust systems with the ability to adapt themselves to system perturbations and failures and thus may yield highly robust systems with the ability to scale freely to almost any size. In this position paper the authors describe the well-established process of use case based derivation of non-functional requirements in energy systems and propose a mapping strategy for aligning properties of self-organizing systems with the ICT- and automation system requirements. It is the strong belief of the authors that such a mapping will be a key factor in creating acceptance of and establishing self-organization in the domain of electrical energy systems.