The data centre market is entering a period of unprecedented transition. With this shift comes a number of significant – and perhaps surprising – changes.
Smaller, enterprise-owned data centres are shrinking in size. At the other end of the scale, mega-facilities continue to grow. What are the dynamics behind this evolution, and how can you capitalise on it to position your business for success?
Big gets bigger: the rise of the mega data centre
Today’s large data centres, typically owned by industry and cloud-provider heavyweights, see nothing but growth on the horizon. They’re becoming “uber-environments”, some sprawling over hundreds of thousands of square metres. According to Kevin Leahy, Dimension Data’s Group General Manager for the Data Centre Business Unit, these mega-facilities are expanding because they’re either absorbing the capacity transferred to them from companies looking to exploit the cloud model, or simply as a result of the nature and complexity of the businesses they serve. “The mega data centre business model focuses on efficiency,” he explains. “Their operators invest in the most advanced technologies for power and cooling, so their energy consumption per unit of computing power is much lower than you could achieve in your own data centre. eBay, for example, has invested in “fuel cells” − devices that turn fuel such as natural gas into electricity through a chemical reaction.”
Mega data centres are typically built to the highest tier 3 availability standards and are located at the lowest latency points in the network. This gives them an advantage when it comes to availability and network performance. They can also scale impressively, the cost of adding a few thousand users to a particular service or application is minimal.
Small gets smaller: the evolution of the business response centre
At the other end of the scale, smaller, enterprise-owned data centres are contracting and evolving into more efficient and agile business response centres. Leahy explains this journey of transformation: “Data centres have typically been capex-intensive, 20-year investments. Over the last few decades, most organisations have found themselves caught in a cycle of on-going expansion every few years, to relieve the pressure on existing facilities that are starting to bulge at the seams. As a result, the cost of owning and operating traditional data centre infrastructures has consumed the greatest portion of the average ICT budget.”
This cycle has abruptly – and permanently – been brought to a halt by a number of elements, including cloud, automation, and technology advancements.
Cloud is the master game changer. By moving certain workloads to the cloud, you can deliver many of your business services to users on an opex, pay-per-use basis, as opposed to owning and operating your own infrastructure. For example, by moving workloads such as email or backup and disaster recovery – which can account for up to 30–40% of your overall data centre capacity – to the cloud, you can extend the life of your data centre significantly and eliminate the need to invest in additional facilities. The cost, scalability, and agility benefits of cloud are widely recognised and have fuelled its rapid move into the mainstream.
But that’s not to say there’s no longer a place for on-premise facilities. The technologies data centres use in order to operate have advanced significantly over the last decade. This means it’s now possible to optimise the performance, consumption, and adoption of your existing infrastructure, and still keep pace with on-going data growth and the ever-increasing demand for capacity. According to Leahy, many businesses have over planned their data centres. “Most companies” existing data centre capacity exceeds their projected needs over the next three to five years. If you can extract more value from the technologies you have in place, it’s entirely possible to shrink your data centre footprint.”
New tools, new ways of thinking
Converged infrastructures and automation are key tools at the disposal of IT leaders looking to breathe new life into their on-premise data centre environments. Says Leahy: “If you’ve automated your environment, you no longer need people manually operating your facility. If you don’t have people there, you can run it at a higher temperature. Research indicates that increasing the temperature at by just five degrees can result in energy cost savings of up to 10%.”
Users – and their locations and work styles – are other factors changing IT leaders” thinking about modern enterprise data centre approaches. Today, many employees aren’t based at company headquarters. Many are mobile workers; others are based at home, satellite, or regional offices. Businesses used to build data centres in their office building to be close to their employees, so that employees benefited from optimal application performance. Now, with modern applications being standardised on new technologies, their performance doesn’t depend on being located in the same building, or indeed on the same continent, as the people using them. You also have more connectivity options: you can move your data centres to locations where the network services can guarantee the same, if not better, service quality than before.
Leahy argues that technology advancements are allowing businesses to correct other flaws with the traditional data centre facility location model. “From a cost and efficiency perspective, there’s never been a good reason to station your data centre in your offices – these buildings don’t have the appropriate power and cooling, and air distribution facilities. Then there’s the question of disaster recovery – another weak point in the old model. Having your servers in the basement of the building makes them vulnerable, for example in the event of flooding.
Consider co-location
According to Leahy there’s another option that should be considered as part of a next-generation data centre strategy: co-location. “Co-location involves placing your assets in someone else’s data centre, where they’re in close proximity to the cloud,” he explains. “The co-location provider is responsible for providing infrastructure systems, and for monitoring, power supply, and temperature control. You retain control of your servers, their accessibility, and security.”
Co-location can address one of the biggest concerns in moving to the cloud: latency between your data centre and the cloud. Co-location providers often have to cross-connect to the cloud provider’s infrastructure, so you benefit from a low-cost, high-performance connection between your assets and the cloud infrastructure.
Transform to better perform – steps to success
* Extend the life of your data centre by making better use of technology advancements.
* Take advantage of cloud and automation.
* Bring the cloud and your data centre together to achieve optimal performance by: optimising the network between your data centre and the cloud; and uniting your data centre and the cloud in the same facility, via co-location.