[Paper Review] Virtualization Implementation Model for Cost Effective & Efficient Data Centers

Mueen Uddin, Azizah Abdul Rahman|arXiv (Cornell University)|Jan 1, 2011

Cloud Computing and Resource Management2 references7 citations

TL;DR

This paper proposes a five-step virtualization implementation model to enhance energy efficiency and reduce costs in data centers by consolidating underutilized servers through virtualization. The model guides data center managers through assessment, planning, deployment, optimization, and management, resulting in reduced energy consumption, lower carbon emissions, and improved resource utilization with minimal downtime and enhanced resilience.

ABSTRACT

Data centers form a key part of the infrastructure upon which a variety of information technology services are built.They provide the capabilities of centralized repository for storage, management, networking and dissemination of data.With the rapid increase in the capacity and size of data centers, there is a continuous increase in the demand for energy consumption.These data centers not only consume a tremendous amount of energy but are riddled with IT inefficiencies.Data center are plagued with thousands of servers as major components.These servers consume huge energy without performing useful work.In an average server environment, 30% of the servers are "dead" only consuming energy, without being properly utilized.This paper proposes a five step model using an emerging technology called virtualization to achieve energy efficient data centers.The proposed model helps Data Center managers to properly implement virtualization technology in their data centers to make them green and energy efficient so as to ensure that IT infrastructure contributes as little as possible to the emission of greenhouse gases, and helps to regain power and cooling capacity, recapture resilience and dramatically reducing energy costs and total cost of ownership.

Motivation & Objective

• Address the growing energy consumption and carbon emissions from data centers due to underutilized servers.
• Provide a structured framework for implementing virtualization to improve resource efficiency and reduce operational costs.
• Enable data center managers to transition to green IT by minimizing physical server sprawl and optimizing power and cooling usage.
• Ensure high availability and resilience during virtualization by addressing single points of failure through redundancy and clustering.
• Support decision-making on virtualization adoption through cost-benefit analysis, ROI estimation, and SLA compliance.

Proposed method

• Categorize existing servers and applications into resource pools based on workload type, utilization, and criticality.
• Assess server workloads for virtualization potential, focusing on underutilized or idle servers (e.g., 30% idle in typical environments).
• Deploy virtualization using hypervisors (e.g., VMware, Microsoft Hyper-V, Citrix) to consolidate multiple workloads onto fewer physical machines.
• Implement redundancy and clustering (e.g., VMware HA, Microsoft Failover Clustering) to eliminate single points of failure in virtualized environments.
• Establish a two-layer management model: Resource Pool (RP) for infrastructure abstraction and Virtual Service Offering (VSO) for service delivery and SLA enforcement.
• Use offline migration (e.g., physical-to-virtual or live migration) to maintain service continuity and meet SLAs during transition.

Experimental results

Research questions

• RQ1 How can virtualization be systematically implemented in data centers to reduce energy consumption and carbon emissions?
• RQ2 What are the key technical and managerial steps required to ensure successful virtualization adoption without disrupting service delivery?
• RQ3 How can virtualization mitigate the risks of single point of failure when consolidating multiple workloads onto fewer physical servers?
• RQ4 What are the measurable cost and energy efficiency improvements achievable through server consolidation using virtualization?
• RQ5 How can data center managers balance ROI, security, and service level agreements during virtualization planning and deployment?

Key findings

• Virtualization reduces energy consumption by consolidating underutilized servers, with 30% of servers typically idle and consuming power without productive work.
• The proposed five-step model enables data centers to reduce total cost of ownership (TCO) by minimizing hardware, power, and cooling expenses.
• Server consolidation through virtualization can significantly lower carbon footprints by decreasing electricity use, which is critical given that data centers consumed 1.5% of total U.S. electricity in 2007.
• The model supports high availability via clustering and redundancy, reducing downtime risks associated with single physical servers handling multiple virtual workloads.
• Virtualized environments improve agility, simplify management, and enable easier backup, recovery, and migration of applications across platforms.
• Proper implementation requires assessing infrastructure readiness for increased power and cooling density, especially in high-availability virtualized setups.