Andy Connor, EMEA channel director, Subzero Engineering looks at the importance of enclosures.
Underpinning this fast-moving digital world is not just the IT software and hardware, but the foundational layer of the data centre, At the very least, the data centres unable to meet digital demands because of ageing, inflexible infrastructure, are in real danger of losing business to competitors. The challenge right now for the data centre industry is ensuring that its facilities can match the speed of innovation, even including their enclosure choices.
The majority of existing legacy data centres consist of large data halls, with much, if not all of the supporting M&E infrastructure attached to the walls and ceiling of the building, and/or under a raised floor. The cabinets may well have remained in the same place since day one. In simple terms, much of the infrastructure and enclosures are fixed, inflexible, inefficient and difficult to move. Modular data centre technology has, therefore, been developed to address these problems. Many modular solutions often consist of a containerized system, which comes pre-populated with racks, cabling, power and cooling infrastructure.
Modular, containerized enclosures will certainly go some way to addressing the digital data centre requirements, but it’s essential to focus on the customer requirement and recognize that some level of customization may be required. Another approach may be to base a modular data centre modernization program on a structural steel frame solution, which is secured to the data centre floor. This allows the owner/operator to make their own choice of modular UPS, cooling, conveyance and cabling components, all of which can be installed as required, along with the IT cabinets and the necessary containment – side panels, doors and ceiling.
Data centre infrastructure vendors who produce, manage, and build traditional data centre infrastructure components cannot keep up with the hyperscale data centre industry's pace of play. The tech firms that are contracting out these hyperscale facilities are not only some of the largest in the world, but they are also, in fact, some of the largest in history. To follow the age-old adage that “necessity breeds innovation” the necessity for these companies to go faster to support data consumption that is already breaking existing networks has spawned innovation that contradicts typical data centre construction methodology.
When utilizing traditional data centre construction an entire data hall’s middle infrastructure would have to be built out before a single rack could be supported. This created a CapEx-heavy environment that added extreme pressure to hyperscale project construction timelines and budgets. The ability to break down and simplify a middle infrastructure solution into a “pod” that could be inserted into a facility on an as-needed basis, would allow for a much lower capital expenditure to rack ratio. Then another of these smaller, simpler middle infrastructure “pods” could be stood up at a later date when needed with minimal or no impact to the previously installed “pod.”
Many traditional methods for supporting data centre infrastructure including containment, power distribution, and cable routing can be costly and time-consuming if not properly designed or specified. They often require structural ceilings, underfloor pathways, and a building that can support the entire weight of the environment. Specifying the load of the system, therefore, is crucial and will influence how the system is designed and configured.
The nature of the IT industry means that, in terms of feeds and speeds, there will always be a part of the infrastructure that is, relatively speaking, something of a bottleneck. Compute, storage, and network components very rarely align. However, there’s no doubt that the digital age is being driven by the fact that all three of these technologies continue to make significant performance improvements, which continue to increase.