From the Lab

Taking the Fear Out of Data Center Migration

Data centers are the heart of any large-scale enterprise. And because they house the information and applications that make business possible, it’s hard to imagine disrupting one voluntarily. Sometimes, however, a company outgrows its available space, or decides that projected performance issues are unsolvable within the existing environment. When that happens, a data center migration is in order.

Unfortunately, migrating from one data center to another is often fraught with concerns about loss of data and productivity, and the potential for spiraling costs. With that in mind, Keystone NAP is aiming to take some of the fear out of the process. As with so many things in life, completing a successful migration all comes down to having an effective plan in place.

Here are the steps we follow – a simple rubric for ensuring that the process runs smoothly, and that the end result is a sustainable foundation for higher performance. This is a process any company can follow on its own. Or Keystone NAP can help.

1. Inventory – Closely examine and document existing data center operations. At the application layer, this includes understanding issues like how different software programs draw down resources and how any associated data for those programs is stored and secured. At the physical layer, an inventory should include everything from recording the power profile of individual servers to calculating network bandwidth needs and documenting overall redundancy, and security requirements.

2. Assessment – Once the inventory process is complete, assess what the end state of a migration should look like. What are the baseline performance metrics that need to be met? Are there ways to improve efficiency? How much room is needed for potential future growth? A data center migration is a chance to do some housecleaning. The end state doesn’t have to look like the beginning. It can and should look a lot better.

3. Action Plan – Develop a roadmap for how the migration will occur. In a perfect scenario, it will be possible to transfer the existing data center workload to a redundant site while the migration takes place. If that’s not possible, determine how the migration needs to be sequenced to minimize disruption. How do uptime requirements differ for different applications, and how should that inform timing? At the physical layer, plan out any equipment that needs to be purchased or upgraded. Factor those details into the larger timeline.

4. Testing and Validation – Based on the performance metrics outlined during the assessment process, test and validate each system as it comes online in the new data center environment. For specific software applications, tests should be jointly developed with business users to make sure that performance in real-world situations meets or exceeds expectations. For ongoing quality control, institute processes for regular system monitoring and set up check points for re-evaluating baseline performance requirements in the future.

If there are major benefits to be gained from moving to a new data center location – and there often are – then the migration process itself shouldn’t be a barrier to achieving those benefits. With careful planning, migration can be smooth, straightforward, and largely worry-free.