Change for the Better

I'm a geek. Yes, I admit it. I've had a lifelong love for technology. I love tinkering with technology and making it do unexpected things. I especially love a challenge -- I love being stretched in new directions and gaining new understanding as I work through some technological hurdle. I'm willing to bet that many of my readers are the same way.

You might say that geeks are naturally disposed toward virtualization. In the early years of virtualization's adoption, you would've been correct. It was the geeks who catapulted virtualization into the limelight after years of running multiple operating systems on their computers with products like VMware Workstation or Microsoft Virtual PC.

In this case, though, my point is: When it comes to virtualization today, the real challenge isn't the technology. That's right. The real challenge isn't virtualization or how to put virtualization to work in your data center. No, the real challenge today is how to deal with the non-technological changes that result from adopting virtualization. To what non-technological changes am I referring? I'm talking about all the ways you'll want to change your organization so that you can derive the greatest benefit from the adoption of virtualization. I believe that in many cases, a failure to change your organization will inhibit the overall value of the virtualization initiative.

Let's look at some examples. How does your organization handle server provisioning today? In many organizations, the hurdles around the acquisition of a physical server -- the expense of the server hardware; the time required by the IT department to install the server and the associated software; the work needed by the networking and the storage teams to configure and provision resources -- keep server growth limited. What happens when those hurdles are no longer present? What happens when tuning up a new server is as simple as cloning a template, and your organization can have a new server up and running in 15 minutes? Some changes are required in these situations, or your virtualization infrastructure will be overrun.

This illustration naturally leads the discussion into another example. How will your organization handle capacity planning once you've virtualized? Before virtualization, when a department or internal business unit needed to deploy a new application, you were deploying capacity for that application when you deployed the application itself. That all changes when you've virtualized your data center. Now you must change the way you approach this problem to ensure you don't run out of capacity and impact the service levels you're able to provide to your organization.

Service levels are yet another area that changes with virtualization. How do you ensure that applications are meeting the service levels you've agreed to provide? Virtualization is about inserting layers of abstraction between computing resources, and those layers of abstraction can make it more difficult to determine if performance problems exist and, if so, where the bottlenecks reside. Your existing tools might not work as well in monitoring the performance and efficiency of your data center after you've introduced virtualization.

I could go on and list more examples, but I think you're getting the picture.

As a geek, I'm of the opinion that more technology always fixes things. OK, maybe not in every case, but in this case VMware and other vendors are stepping up with technology that will help address some of these non-technological challenges associated with virtualization. A number of vendors are now offering products that provide frameworks or workflows around the virtual machine (VM) lifecycle, so that companies don't inadvertently cause "VM sprawl." Similarly, organizations are offering products for capacity planning and trend analysis so that you can get -- and keep -- a handle on resource allocation and utilization. Finally, more and more products are appearing that are expressly designed to work within virtualized data centers to provide the right metrics and correlate those metrics together for an accurate picture of how your virtual infrastructure is performing.

As you embrace virtualization within your data center -- either for the very first time, or perhaps as you expand the use of virtualization to include more applications than you had in the past -- don't forget to address the non-technological challenges as well.

About the Author

Scott Lowe is a virtualization architect for ePlus Technology in Herndon, Va., and author of "Mastering VMware vSphere 4" (Sybex, 2009).


Subscribe on YouTube