Enterprises are taking a closer look at transforming their data centers to allow them to better handle the demands of disruptive information technology (IT) shifts and other key demands on data centers. According to a recent survey of more than 320 IT pros by QuinStreet Enterprise, 61 percent of respondents are modernizing their data centers and feel it is a priority.
In a recent post here, I outlined four key triggers behind data center transformation. Disruptive technologies such as Cloud computing and Big Data are one trigger, but so are priorities such as regulatory compliance and curbing data center energy costs. Once we recognize these triggers and accept the need to transform, the question becomes, how do you best approach a data center transformation?
No one would build a major new data center without careful planning. Requirements would be gathered, expected loads would be calculated, site analysis would take place, culminating in the use of facility design software to model the details of the to-be data center. Similarly, when you are looking at the need to transform an existing data center or multiple data centers to meet growth demands, you also need to plan. You need a methodology.
At Schneider Electric, we see data center transformation as a process that can be addressed under a methodology that spans the following phases:
- Data center transformation strategy. This addresses “big picture” questions such as whether you want to outsource some load to a colocation data center or Cloud services, add capacity to an existing facility, or build a new facility. Enterprises often need help modeling the costs and paybacks of the alternatives including the energy costs. This is the stage the outcomes are well defined.
- Audit and review the current critical infrastructure. This step focuses on assessing the capabilities and useful life of what you have in place against emerging business requirements and outcomes. Here is where a detailed energy assessment, assessment of IT and data center physical infrastructure and a holistic operations assessment are needed to gauge what you can build on or what needs to be upgraded, replaced, or expanded upon.
- Critical infrastructure design. If new capacity or facilities are needed, a design phase needs to take place, taking into account key parameters such as criticality, capacity, growth plan, efficiency, density, and budget. Today, services partners for this step not only need experience with a more traditional (physical) data center environment, but also virtual and cloud data center models, or a mix of all three. At this stage the end state management and monitoring design is defined.
- Data center transformation and transition. This is the actual build phase of the to-be infrastructure. It is vital to have a data center partner who can bring in industry leading providers across categories such as IT hardware, networking, and IT services.
- Data center operations. The startup phase for expanded or new operations needs to be supported by thorough planning and training. To achieve a more proactive data center, it’s likely that use of advanced analytics and tools such as data center infrastructure management (DCIM) software will be part of the mix. In some cases, the owner/operator may want to opt for data science services that employ Big Data analytics to optimize operations.
- Continuous improvement. To transform over the long term, you need a foundation for continuous improvement and a proactive management environment. Dashboards and DCIM can help, as well as advisory services from a firm steeped in guidelines for green data centers and other best practices, such as the Data Center Maturity Model from the Green Grid.
Unfortunately, there is no single technology or product that is going to transform every data center. To succeed at transformation, the most important thing is to follow a methodology. Secondarily, it helps to have a partner with deep and broad skill sets. A DCIM specialist isn’t the answer, nor is a facility designer or architect. These skill sets might be called upon within data center transformation projects, but you need a data center transformation partner whose expertise spans the multiple domains involved in the planning, design, building, operation, and improvement of data centers.
Remember, data center transformation is more than an upgrade. Yes, you might be after “x” amount of new capacity, or you might decide it’s important to establish a hybrid environment that can tap into Cloud capacity, but beyond that initial “spec,” you also want a data center that is responsive and agile enough to accommodate new requirements. That’s why the last step in the transformation methodology—continuous improvement—is so important.
In this age when the way that companies sell products, market to customers, and manage assets is highly digitized, and thus generating massive new IT loads, it’s necessary to transform data centers into engines for growth. Incremental performance improvements in existing data centers is fine, but to support growth today, more and more companies are going to have to transform their data centers.