There is a clear economic case for public cloud adoption, and every IT shop has heard it by now: You can benefit from a pay-as-you-go model with near infinite elasticity for applications with varying demands. Nearly half of large enterprises will get there by the end of 2017, according to Gartner.
But what about the logistics and costs of migration? What about the challenges of incorporating public cloud services into existing IT operations? What about figuring out how disaster recovery and business continuity will work in a hybrid cloud landscape? Enterprises may struggle to answer these questions with traditional approaches and solutions, but the issues can be overcome. Follow these best practices to reduce the risk and increase your potential for early success:
Head to cloud school.
You need to learn the basics. So start by reading about all materials available to you, such as cloud migration tutorials and getting-started guides. Make sure your team's experience includes a range of hands-on learning opportunities. Once you wrap your head around the concepts, gain valuable experience by starting with the migration of a single workload -- and start with the easier ones before you tackle more complicated multi-tier applications (even if that's your ultimate destination).
Look for the technical support that's right for YOU.
When you're weighing public cloud provider choices, focus on technical support early in the process -- not after you make your selection. Ask the hard questions about who will be on your team, what skills those people have, what services are included in the SLAs and what will happen if you need to escalate a concern. A big part of these conversations should focus on cost; you don't want to lose all the economic benefits of cloud adoption to unexpected service and support fees.
Map your cloud plans for each application.
While it's great to start with a single workload or project, don't head for the cloud without first establishing an overall plan. Consider which applications you want to move to the cloud permanently and which ones might be there only occasionally when demand exceeds on-premise resources. Be realistic about which mission-critical, complex applications are better suited to stay on premise -- at least in the initial stages as you begin to shift certain workloads to the cloud.
Don't go all in all at once.
Be realistic about how many virtual machines (VMs) you can import in the first migration phase. Keep the number manageable, and remember that you need to build in time for the users to test the applications to ensure migration efforts will be successful. Involve your friendliest users in that testing process at first, and then broaden your test to include your entire user audience. Look for any broken dependencies, missing configurations and other hidden problems. When you're satisfied that the migration was successful, move onto the next group of VMs -- with increasing complexity or criticality.
Let people know what's coming.
When IT rolls out any new project, many of the pain points users complain about are either caused or exacerbated by a lack of communication. Talk to users about what's coming, what problems might surface and why it benefits the enterprise (and them) to make this transition. Recruit some alpha and beta groups to help test drive your public cloud initiatives. Listen to what they have to say about any bumps in the road before going live with the rest of the company.
Extending on-premise IT infrastructure to the public cloud for a hybrid IT computing model makes a lot of sense, and it's easy to see why so many enterprises are pursuing plans to migrate. To make sure unexpected expenses and problems don't erode the benefits, educate yourself, plan ahead, partner with skilled service providers and vendors, and communicate openly and realistically with end users.