It seems like cloud is all anyone can talk about these days. But virtualization is an important step toward developing a private cloud strategy. If you’ve already virtualized part of your IT infrastructure, you’re probably closer to private cloud computing than you think.
Cloud’s getting all the attention these days because of the benefits of cloud computing and the inherent limitations of virtualization. Virtualization concerns itself with the virtual machine (VM), how well that VM performs and what can be done once a physical server is virtualized. Private cloud computing takes a wider view, focusing not on the VM itself but instead on the entire infrastructure in which VMs are hosted.
Why develop a private cloud strategy?
Looking past the hype, are the benefits of cloud computing worth attaining? Most assuredly. For a solid private cloud strategy, start by embracing the concept of the virtual data center. That approach will introduce the benefits of cloud computing -- further-optimized resources, better ability to deliver services and improved visibility into resource usage -- to your IT infrastructure.
So how do you get there? The following four steps can help you develop a private cloud strategy and reap the benefits of cloud computing.
Step 1: Recognize what private cloud computing is
The IT industry has done itself a disservice by referring to this whole concept as “cloud.” Too many people see vaporware in what is really a powerful new mindset for managing IT workloads. At
Self service is the most important functionality of that resource pool. You can set up a cloud of resources, dole out a portion to specific individuals, teams or projects, and allow them to use the resources in whatever way they see fit. More and more self-service tools are becoming available, some from virtualization platform vendors and others from third parties or customized from public cloud vendors.
If you define private cloud computing using VMware’s suite of management tools, it’s little more than a cluster of hosts that have VMware High Availability and Distributed Resource Scheduler enabled, integrated with the self-service tools mentioned above. With these assets, along with the storage and networking that accompanies them, you’re able to flexibly create VMs up to the level of your supply of physical resources.
So with what you’ve already got today, you’re well on your way to developing a complete private
Step 2: Recognize the private cloud computing components you don’t have
Many IT professionals don’t actively manage system performance, even after virtualizing. But to get the benefits of cloud computing, performance monitoring should be a critical part of your private cloud strategy.
Private cloud computing represents an abstraction of the entire data center. That abstraction consolidates hardware into a set of numbers that measure capacity. Network, storage, processing and memory all are abstracted into numbers that quantify resource supply and demand.
You see evidence of this abstraction today. For example, pull up the VMware vCenter Client’s Virtual Machines tab for a cluster and you’ll find a long list of VMs with their processing and memory demand values. Advanced tools such as VMware vCloud Director and System Center Virtual Machine Manager 2012 (currently in beta) bring further visualization of these resource values.
Because of the abstraction, it’s important to embrace performance and capacity management at a
data-center level when adopting private cloud computing. You’ll need more tools than what your
virtualization platform alone can provide, but you’ll also need an evolved approach to IT
resource management that regards your assets as contributions to a whole.
Step 3: Cease and desist the virtual white boxing
There’s a new class of hardware now available from major manufacturers: converged infrastructure. This hardware is modular, making it easy to add computing power, storage or networking throughput by simply snapping in additional capacity. And manufacturers bundle in management tools to govern this hardware for virtualization.
That handshake between hardware manufacturer and virtualization platform is the real linchpin of private cloud computing. It’s the connection that enables admins to further optimize how VMs consume resources in their private cloud strategy. Storage works with servers, which communicate across networks, which combine to create a seamless experience for virtual workloads.
These technologies are available today, but you won’t always be able to make them part of your cloud strategy overnight. With hardware refresh cycles the way they are, it may take a few years for this new equipment to make its way into your data center. What you can do for now is plan for its arrival.
Here’s why: You learned a decade ago that building white-box servers from scratch might be loads
of fun, but these dissimilar servers and their configurations grow unmanageable as the
infrastructure scales upward. Unless you begin planning for converged infrastructure now, you’re
doomed to relearn a similar lesson with your entire data center, and your homemade virtualization
hardware won’t scale to meet your needs.
Step 4: Right-size services to their delivery platform
To really achieve the benefits of cloud computing, you need to get over your fears about security and loss of control.
Private cloud computing is the first step toward a future where IT services can be flexibly hosted wherever they make sense. Sometimes it makes sense to host those services in your local data center. Other times, it makes more sense to let someone else do the hosting in a public cloud.
Bridging these two methods are an evolving series of technologies that secure the connection,
protect the information and create the hybrid cloud experience. Both your cloud services vendor and
your virtualization platform vendor can now share with you the current and future vision for these
products, because they are today -- finally -- products that you can actually see and
Private cloud computing: Closer than you think
Not long ago, there was the notion that a thin wafer of plastic could never be a secure mechanism for purchasing goods and services. The credit card, with little more than a series of numbers and a magnetic strip, was a commerce vehicle that people weren’t ready to trust.
Today, we have a level of trust that’s become so engrained we don’t think twice when we purchase something. It’s a perfect metaphor for the evolution of trust we’re seeing in IT today. Private cloud computing is one part of that trust, and its bigger brother, public cloud computing, isn’t far behind.
This was first published in June 2011