To build a successful cloud architecture, your people, processes, and technology must all work in unison. Don’t shoehorn people and process into the technology, make it support your people and processes. How do they work together? What are your end goals? Are you trying to optimize for speed, or reduce costs? The answers need to guide your architecture. Map them out so you can identify and solve for inefficiencies—and build tools that work for the people that will use them.
Inevitably, you’re going to run into issues as you modernize your infrastructure. When you decide to move to a private cloud, you need to assess all the contributing factors because technology is probably only one piece of a complex, interdependent puzzle. That means paying attention to people, processes, and technology.
Give power to the people
Your first job is to understand precisely what the people who will be using your private cloud need. Some of your DevOps people might be best served by relatively simple tools that don’t require them to make too many decisions. For example, they might just want to request a server without worrying about the exact CPU or storage. Others, like NetOps people, will need more control over details, such as being able to choose among networking and integration options. Still others may require special services, such the ability to monitor servers and grab logs on demand.
How will all these different user requirements work together? The self-service nature of private clouds means that you, as an architect, have an opportunity to streamline operations by presenting the right services to the right users. But doing that effectively means taking the time to truly understand how each category of user is best served. If users are struggling, check to be sure you got the mapping right—and if not, you have identified a problem to address.
Processes are another matter. The technology that you purchase to enable the abstraction we call the private cloud doesn’t automatically provision or de-provision things. It doesn’t do workflow processing. It doesn’t do approvals or auditing or any of the other things you need to do to keep your IT infrastructure humming along. None of this comes out of the box. A private cloud simply gives you basic constructs in compute, network, and storage. You have to figure the rest of it out.
For example, private cloud doesn’t come with a utility billing structure in place. This is a business process concern, not a technology one. If you are struggling with resource allocation—perhaps too many resources are being consumed by low-priority projects, leaving high priorities underserved—then examine whether you have effectively replicated the appropriate controls in your new infrastructure. How do you get users to pay for what they use? What’s the approval process? What if a department or user runs out of money or credit? This kind of resource allocation problem was fixed long ago in the standard data center, but now you have to figure it all out again with private cloud.
Be sure to map the technology to your users and processes. Your private cloud should be about technology in service of your people and processes, not the other way around.
Plan for rapid technical change
Understand that much of what you invest in today is going to be inevitably weighed down by entropy in less than three years. And the pace of change is accelerating. As your own needs evolve, you’ll want to make changes. That’s why you choose flexible, extensible, and adaptable infrastructure components when building your private cloud. Get as close to plug and play as you can get in this highly chaotic world. That way you’ll be able to swiftly address the people and process issues that will inevitably arise.
Alice LaPlante is an award-winning and best-selling American author of numerous books, including A Circle of Wives and The New York Times bestseller Turn of Mind, which was the winner of the Wellcome Trust’s Book Prize and a B&N Discover Award finalist. She was a Wallace Stegner Fellow and a Jones Lecturer at Stanford University, and taught creative writing at both Stanford and San Francisco State University. She has written for Forbes ASAP, BusinessWeek, ComputerWorld, InformationWeek, and Discover. Her corporate clients include some of the best-known brands in the technology industry, including IBM, Microsoft, Oracle, Symantec, Deloitte, and HP.