Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

SMART ENERGY MANAGEMENT IN DATA CENTERS

IP.com Disclosure Number: IPCOM000246061D
Publication Date: 2016-Apr-29
Document File: 9 page(s) / 67K

Publishing Venue

The IP.com Prior Art Database

Related People

Mehiar Dabbagh: AUTHOR [+2]

Abstract

Techniques are provided for reducing peak power consumption for data centers through the use of power storage devices. These techniques may determine a current charge level for each of a plurality of power storage devices within a data center. A predicted future power consumption of the data center may be determined over a future interval of time. A predicted maximum amount of power drawn by the data center could be determined in each of a plurality of time slots within the future interval of time. Additionally, a scheme is provided to generate a schedule that provides an optimal usage of the plurality of power storage devices that minimizes the maximum amount of power drawn across all of the plurality of time slots within the future interval of time. The computing devices in the data center can then be scheduled to operate using the plurality of power storage devices according to the generated schedule.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 15% of the total text.

Page 01 of 9

SMART ENERGY MANAGEMENT IN DATA CENTERS

AUTHORS:

Mehiar Dabbagh Ammar Rayes

CISCO SYSTEMS, INC.

ABSTRACT

    Techniques are provided for reducing peak power consumption for data centers through the use of power storage devices. These techniques may determine a current charge level for each of a plurality of power storage devices within a data center. A predicted future power consumption of the data center may be determined over a future interval of time. A predicted maximum amount of power drawn by the data center could be determined in each of a plurality of time slots within the future interval of time. Additionally, a scheme is provided to generate a schedule that provides an optimal usage of the plurality of power storage devices that minimizes the maximum amount of power drawn across all of the plurality of time slots within the future interval of time. The computing devices in the data center can then be scheduled to operate using the plurality of power storage devices according to the generated schedule.

DETAILED DESCRIPTION

    Cloud computing has become a popular approach for obtaining access to (sometimes large-scale) computing resources. Cloud computing allows users to build virtualized data centers which include compute resources, networking resources, applications, and storage resources without having to build or maintain a physical computing infrastructure. The virtualized data center may provide a user with a segmented virtual network located in the cloud, typically alongside virtualized data centers of other users. Such a virtualized data center may be rapidly scaled up (or down) according to the computing needs of a given user without the need to maintain excess computing capacity between peak demand periods. For example, an online retailer can scale a virtualized data center to meet increased demand during the holiday shopping

Copyright 2016 Cisco Systems, Inc.

1


Page 02 of 9

season without having to maintain the underlying physical computing infrastructure used

to provide the retailer's online presence.

    Oftentimes, a cloud-computing environment is created using multiple data centers, with each data center providing various computing resources to the cloud. Such data centers are frequently located in different geographical locations. Furthermore, the resources that each data center provides to the cloud may differ. For example, a first data center may provide higher performance computing resources than a second data center, or may provide fast network access to particular computing resources that are not provided at all by the second data center. Additionally, the workloads of the computing resources provided by each of the data centers may differ as well. For instance, while the computing resources of the first data center may be operating at 90% capacity, the computing resources of the second data center may only be operating at 20% capacity.

    When deploying virtual workloads within a data center, conventional systems...