Energy and performance management in large data centers: A queuing theory perspective

Abstract

With the increasing popularity of Internet services and cloud computing, both energy efficiency and performance guarantees are major concerns for data center operators. However, only a few of previous works elaborate the dynamic system working procedure while using specific mathematical methods to study on the relationship between the two characters. This paper addresses how to guarantee performance requirement in terms of waiting time so as to minimize power consumption by means of dynamic management policies that switch on/off a certain group of servers. We develop a mathematical model using queuing theory to determine activation thresholds of servers. Our scheme provides an analytical approach for studying the power-performance tradeoff in large data centers.

An Energy-Efficient Task Scheduling Mechanism with Switching On/Sleep Mode of Servers in Virtualized Cloud Data Centers

“…In [23], Liao et al minimized the power consumption by switching on/off a certain group of servers according to whether the current queue size reached the threshold. ey introduced two activation thresholds and modelled the data center as an M/M/n + m1 + m2 queueing system.…”

An experiment-driven energy consumption model for virtual machine management systems

“…The virtualization proposes to improve resource utilization through dividing one physical server into multiple isolated virtual machines. Resources sharing makes complex to leverage experiences from previous efforts on energy estimation in single host [8,9,10,11,12,13] and on energy estimation in cluster models for data centers [14,15].…”

Stabilizing the virtual response time in single-server processor sharing queues with slowly time-varying arrival rates

“…Modern data centers consume tremendous amounts of energy to supply networking, computing, and storage services to global IT companies. Concerns about energy consumption have prompted researchers to explore operational methods that maximize energy efficiency and satisfy a certain level of quality of service (QoS), [2,3,4]. QoS can be achieved by adding constraints that impose upper bounds for response time-related metrics, e.g., the mean virtual response time and the tail probability of the response time.…”

“…Binding the QoS-related constraints implies that the metrics are maintained as a constant value, and suggests the need to investigate the stabilization of response times. Although some proposed methodologies [2,3,4] assume the stationarity of data traffic arrival processes, nonstationary properties, such as time-varying arrival rates from real data [5], make it difficult to analyze queueing system performance.…”