Power and Performance Management of Virtualized Computing Environments Via Lookahead Control

Abstract

There is growing incentive to reduce the power consumed by large-scale data centers that host online services such as banking, retail commerce, and gaming. Virtualization is a promising approach to consolidating multiple online services onto a smaller number of computing resources. A virtualized server environment allows computing resources to be shared among multiple performance-isolated platforms called virtual machines. By dynamically provisioning virtual machines, consolidating the workload, and turning servers on and off as needed, data center operators can maintain the desired quality-of-service (QoS) while achieving higher server utilization and energy efficiency. We implement and validate a dynamic resource provisioning framework for virtualized server environments wherein the provisioning problem is posed as one of sequential optimization under uncertainty and solved using a lookahead control scheme. The proposed approach accounts for the switching costs incurred while provisioning virtual machines and explicitly encodes the corresponding risk in the optimization problem. Experiments using the Trade6 enterprise application show that a server cluster managed by the controller conserves, on average, 22% of the power required by a system without dynamic control while still maintaining QoS goals. Finally, we use trace-based simulations to analyze controller performance on server clusters larger than our testbed, and show how concepts from approximation theory can be used to further reduce the computational burden of controlling large systems.

A Hierarchical Correlation Model for Evaluating Reliability, Performance, and Power Consumption of a Cloud Service

“…Denote this performance metric of the cloud service as γ . When the application has x available VMs serving user requests, from (19) and (26), we can obtain γ (x) as Now, we can implement a MRM to connect reliability with performance. In addition to the steady probability p a (x)(x = 1, 2, .…”

“…Given this scenario, the probability that a new request is discarded by the COS can be obtained as p block = 0.0672 by using (19). Substituting the given parameters in (23) gives the pdf f i−x+1 (t) of the different values of x, which are depicted in Fig.…”

Joint Optimization of Resource Provisioning in Cloud Computing

“…The optimization problem considers bandwidth as a constraint, but does not address routing of bandwidth, and is therefore too simplistic. Live migration is an important ability for balancing load on a network, and [13], [14], and [15] account for bandwidth costs during the migration. A cost function is formulated for the bandwidth of migration whilst maintaining QoS.…”

Adaptive Controller for Dynamic Power and Performance Management in the Virtualized Computing Systems

“…(2). to optimize a new objective, which might integrate several objectives such as the performance, the power, or the balance between different machines, see [7]–[8] etc.. [7] formulates the control problem as a profit maximization problem by integrating the SLA function, which represents the performance, the power consumption, and switching cost. Then two-level control hierarchy is introduced, where one level is a faster control, the other level is a slower control.…”