Software-Defined Cloud Computing: Architectural Elements and Open Challenges

TL;DR

This paper introduces Software-Defined Cloud Computing (SDC) as an approach to automate cloud infrastructure configuration, enhancing QoS management, energy efficiency, and resource utilization through reconfigurable virtualization.

Contribution

It presents an architecture for SDCs focusing on mobile cloud applications and evaluates its potential through two practical use cases.

Findings

SDC enables dynamic reconfiguration of cloud resources.

Improves QoS through bandwidth-aware allocation.

Enhances energy efficiency with optimized VM placement.

Abstract

The variety of existing cloud services creates a challenge for service providers to enforce reasonable Software Level Agreements (SLA) stating the Quality of Service (QoS) and penalties in case QoS is not achieved. To avoid such penalties at the same time that the infrastructure operates with minimum energy and resource wastage, constant monitoring and adaptation of the infrastructure is needed. We refer to Software-Defined Cloud Computing, or simply Software-Defined Clouds (SDC), as an approach for automating the process of optimal cloud configuration by extending virtualization concept to all resources in a data center. An SDC enables easy reconfiguration and adaptation of physical resources in a cloud infrastructure, to better accommodate the demand on QoS through a software that can describe and manage various aspects comprising the cloud environment. In this paper, we present an architecture for SDCs on data centers with emphasis on mobile cloud applications. We present an evaluation, showcasing the potential of SDC in two use cases-QoS-aware bandwidth allocation and bandwidth-aware, energy-efficient VM placement-and discuss the research challenges and opportunities in this emerging area.