Is virtualization a strategy or a tool?

Posted on August 01, 2009

RssImageAltText

By Noemi Greyzdorf

Strategy is defined by Miriam Webster dictionary as the art of devising or employing a plan to achieve a goal. Too often as an industry, we have focused on the tools more than on the desired result, and thus begin to view the tools as a strategy rather than the goals.

As an example, some time ago there were many conversations about information lifecycle management. ILM is an important concept that when adopted properly can deliver significant benefits to the business and IT organizations. Unfortunately, we became more enamored with the tools, such as HSM and storage virtualization, and forgot about what ILM really meant.

The same is occurring again today. We have become so enamored with virtualization that it is no longer viewed as a tool or a means to an end but, rather, the end itself. It is frequently discussed as virtualization strategy. The question though is how can you have a strategy around a tool?

So if virtualization is not a strategy, what is it and how can it help? The real goal in all this must be the drive to achieve greater efficiencies in the data center. This means leveraging existing resources, improving utilization, reducing datacenter footprint, and containing costs. In many ways, these are all interconnected. If you are running out of data center space or power, you are not free to buy more equipment indiscriminately because you might not have any place to put it. More data might require more storage that needs to be managed, but there might not be the personnel available to fulfill all the needs of increasing storage.

If the strategy is to make the environment more efficient, then the tactics of executing on the strategy must consist of processes and systems that create efficiencies.

  1. Assess what you have. Understand how resources are consumed today and by whom, as well as what the actual requirements are for various applications. One aspect of this is information governance.
  2. Establish policies that define how applications are handled, and what the protection and recovery requirements are for these application and their data.
  3. Identify tools that help you achieve the goals set out by the first two steps.
  4. Determine all requirements around deploying, select tools, and then implement.

It is common that when managers go through the first two steps, they find that their infrastructure in many areas is underutilized, takes too long to provision (thus impacting time to market or productivity), or is too costly to maintain for use cases such as test and development or iterative marketing campaigns. On the other hand, there may be many applications that don't get enough necessary resources to perform adequately. If the resources available are under-utilized, one approach to increasing utilization is to virtualize the infrastructure and then allocate just what each application really needs—no more, no less.

The result of virtualization may be consolidation of many systems into one physical device shared among multiple tenants. Additional benefits may include faster provisioning of resources, flexibility in scaling resources, and simplified management.

On the other hand, the use of virtualization may create new challenges, such as troubleshooting from end-to-end through many abstraction layers, managing storage and its performance, and load-balancing applications based on requirements. In order for the use of virtualization to achieve the strategic goals of creating greater efficiencies and reducing costs, the organization must plan its implementation and take the full ecosystem into consideration.

This may seam like an exercise in semantics, but how we talk about something often affects how we end up using it and managing it. Using the right language may help vested parties to see more clearly what the true objectives are and what steps must be taken to achieve them.  


Noemi Grezdorf is a research manager with IDC. She can be contacted at ngreyzdorf@idc.com.

 More InfoStor Issue Articles

Originally published on .

Comment and Contribute
(Maximum characters: 1200). You have
characters left.

InfoStor Article Categories:

SAN - Storage Area Network   Disk Arrays
NAS - Network Attached Storage   Storage Blogs
Storage Management   Archived Issues
Backup and Recovery   Data Storage Archives