Realizing Data Center Efficiency with Virtualization and Dynamic Infrastructure

By Lori MacVittie
published: Friday, August 28 2009

Virtualization is rapidly becoming a fact of life in the day to day operations of data centers across the globe. Whether it's being used for consolidation purposes, as a means to achieve a more dynamic cloud-based architecture, or whether it's required as part of a cloud-based initiative, virtualization is one of the foundational technologies involved in efforts to realize greater data center efficiencies.

While consolidation is often cited as one of the primary means through which value is realized with virtualization - it has an appreciating return because a single physical system can support multiple platforms simultaneously, which lowers the cost of operating expenses related to hardware support - efficiency and agility are just as beneficial to the data center investing in a virtualized infrastructure.

In the past, hardware resources were often underutilized purposefully which meant a lot of money was being wasted by essentially over-provisioning of hardware and software licenses. Virtualization has allowed organizations to maximize the utilization of hardware resources on an as-needed basis, saving time and effort when more capacity is required as "extra" resources can be utilized by applications requiring more capacity without impacting existing applications and their resource requirements. The ability to consume what would otherwise be idle resources is more efficient than traditional means of increasing capacity, and further offers greater agility in deployment and operations as IT staff can respond immediately to unanticipated demand for applications or other resources being served by a virtualized system.

The nature of virtualization also improves the efficiency of the processes used to provision resources as a single virtual machine can be configured once and simply deployed as necessary. This reduces the potential for the introduction of configuration errors or conflicts between applications, operating systems, and ancillary dependent software as once the system has been properly configured it can be provisioned and de-provisioned at will with the certainty that when launched it will be ready to execute.

Benefits of Dynamic Infrastructure

There are two ways in which a dynamic infrastructure improves efficiency in a data center: process efficiency and technological efficiency. Both forms of efficiency are achieved via one of the core properties of a dynamic infrastructure: contextual awareness.

Process efficiency is achieved through automation of the network and application network infrastructure. Operational processes such as provisioning and managing maintenance windows often require manual intervention. The integration and collaborative attributes of a dynamic infrastructure, i.e. a standards-based API providing access to the infrastructure's control plane, allow many of these processes to be automated and integrated into existing management systems. In addition to efficiency, automation of operational processes provides a measure of consistency across time and processes by removing the possibility of human error that can introduce failures.

Process efficiencies require contextual-awareness such that the infrastructure is capable of aiding in decisions regarding when and how to participate. Because the infrastructure can be integrated, it can both provide and take direction from external management and virtual management systems. Thus, it can provide a greater level of detail regarding the need for additional - or fewer - virtual instances of an application based on current conditions in the user, network, and application environment. It may not be efficient, for example, to launch a new instance of an application if the pattern of incoming requests is trending down instead of up. A context-aware dynamic infrastructure can not only track this level of information but provide data in the form of concrete analysis to the systems responsible for process automation.

One example of this is an Infrastructure as a Service provider that has integrated application delivery networking as a component of with its rapid provisioning services.   The system leverages its contextual awareness of the application to make real-time decisions regarding the most efficient and well-performing routing of requests, and passes those requests automatically to the application delivery network devices.  The provider has been able to automate processes that take into account current conditions for each application and request, and that would otherwise have been static and manual.

Technological efficiencies come from the nature of a dynamic infrastructure in its abilities to improve the performance of applications and servers (virtual and physical) through offload technologies. These offload technologies - SSL, compression, TCP session management - require a level of awareness peculiar to dynamic infrastructure devices. Context-awareness provides the means by which the infrastructure understands not only the request, but the context in which it is being sent, i.e. user-specific variables, network conditions, site and application status, and current conditions within the data center. Based on any combination of these variables, or all, a dynamic infrastructure can determine how best to apply its technology to every request and response on an individual basis to ensure the security and performance of the application.

For example, some responses will not benefit from compression. The size, type of content, and network link over which the user will receive the response should all be considered in whether applying compression will be advantageous or not in terms of performance and resource savings. Without an understanding - and awareness - of all those variables it is not possible to make that determination.