Application Infrastructure Virtualization: An Essential Element of a Green Data Center
Application Infrastructure Virtualization: An Essential Element of a Green Data Center
By Samir Mehta
published: Monday, September 08 2008


IBM_20080908.png

Many organizations are increasingly challenged by the growing utilization of their application and hardware servers, the lack of adequate hardware resources and space to house them, the energy needed for application support and the service required for business operations. In order to help meet the demand for IT services, many organizations have made Service Oriented Architecture (SOA) the underpinning technology for their enterprise application strategy. Not surprisingly, increasing complexity has made IT management and administration the leading drivers of IT costs, which are increasing at an alarming rate.

 

Reducing the consumption of key resources such as energy and personnel while improving the utilization of IT hardware and software can yield environmental benefits as well as financial benefits. Virtualization is one means by which organizations can improve the efficiency of the data center and utilization rates for IT resources. Virtualization can also assist organizations in reducing their expenditures for energy. As various components of IT have been added to the virtualization fold, there have been new opportunities for organizations to improve their utilization and energy efficiency.

 

IBM_Quote3_20080908.pngOrganizations that have implemented virtualization in their server farms, data storage, desktops, and networks  are reaping the benefits of simplification, consolidation, and flexibility that are the hallmarks of virtualization technologies. They probably think that they have taken the virtualization road about as far as they can in the data center. Yet, there is another critical aspect of virtualization technology which they may not yet have tapped into - a complementary virtualization approach that can maximize the benefits of an enterprise's SOA strategy.

 

Application infrastructure virtualization (AIV) complements server, storage, and network virtualization and enables businesses to push the boundaries of their IT infrastructures further for greater agility, cost savings, operational efficiency, and manageability.  AIV provides the ability to separate applications from the physical infrastructure that they run on.  Workloads can then be dynamically placed and migrated across a pool of application server resources allowing the infrastructure to dynamically adapt and respond to business needs, and requests to be prioritized and intelligently routed to respond to the most critical applications and users.

It frees the enterprise from a tight coupling between an application and associated application servers. This loose coupling enables open standards based software to intelligently manage and shift workloads according to service policies.  Application infrastructure virtualization enables CIOs and IT administrators to do more with less¾quite literally.  It allows data centers to run applications on any application server in a common resource pool and break down existing silos. Furthermore, administrators can deploy resources quickly and seamlessly during peak periods and in response to unforeseen demand for mission-critical applications. Last, but certainly not least, data administrators can also achieve application response times and service levels that meet strict Service Level Agreements (SLAs)

User client virtualization

Virtualization

within the

data center

Application virtualization

Desktop virtualization

Server virtualization

Storage virtualization

Network virtualization

Application
infrastructure   virtualization

IBM_20080908_1.png

Figure 1. A comprehensive view of virtualization

Reining in Operational and Energy Costs with Application Infrastructure Virtualization

Together with server, storage, and network virtualization, application infrastructure virtualization addresses one of the major causes of rising costs and complexity in the data center¾the tendency for data "silos" to develop in the data center.  A leading analyst firm has revealed that the total cost of ownership for servers continues to rise, despite the fact that spending on servers has actually remained fairly flat. Between 1996 and 2006, the total cost of ownership for servers worldwide rose by more than 150%.

The rising costs are projected to continue and are being driven, first and foremost, by server management and administration costs. The secondary driver behind rising ownership costs for servers is the cost of power and cooling for data centers.  With the cost of electricity and personnel rising, CIOs and IT administrators clearly have their work cut out for them when it comes to containing costs.

Most telling, however, is the hidden cost of the well-established fact that businesses typically buy servers to satisfy peak demand for processing, leaving the vast majority of their data centers' capacity idle most of the time. The utilization rate of the typical server is approximately 15 percent. This low utilization rate means that the cost of unutilized¾not underutilized¾assets in the data center is estimated to be well over $100bn.