Data at the edge
By Robert Linsdell, Managing Director Australia and New Zealand, Vertiv
Monday, 11 September, 2017
Businesses in Australia and New Zealand are changing how they design their data centres, and that means electrical and cooling professionals will need to be across the technology changes as they play their part in the deployment and maintenance of the new age data centres.
Public cloud continues to grow, as do innovations in on-premise data centres, such as hyperconverged infrastructure. But there’s a new kid on the block disrupting the lot — the modular or ‘edge’ data centre.
The edge is hot — by edge, I mean the edge of the network, the mobile devices that organisations now find integral to their day-to-day business; and by hot, I mean more and more data is being created here, at the edge, than ever before.
Enter modular data centres. Businesses across different industries are fast deploying compact, plug-and-play data centres to branch or regional offices to manage the needs for on-site compute and networking or cloud based on premises, largely thanks to the modern mobile nature of working, concerns of data sovereignty and security of data considered too confidential for cloud.
The benefits are many — if you can process data closer to where it’s produced, it reduces latency for time-sensitive applications and eases the constantly increasing bandwidth required to transit back to the cloud or core data centre. This means a better, more reliable user experience at the edge, and no need for large amounts of data to be hauled back to the data centre.
Cloud and co-location providers have observed this change and responded; businesses overseas have begun to establish an array of small to medium-sized modular data centres. In some cases, cloud providers have crafted offerings to arrange their on-premises cloud stack to be housed in enterprise data centres.
Cloud providers know that these data centres can enable customers to build out their ‘blast radius’ — lowering the latency between operating nodes to give them faster access to data. Similarly, co-location providers can host edge applications for enterprise businesses. Both are striving for an enhanced user experience, whether the user is developing a new digital service for another department, processing financial transactions or scanning Reddit for the latest Game of Thrones theories.
The trend towards edge data centres is global, and Australia is no exception. Last year, Frost & Sullivan predicted that modular data centres would play a big part in the wider ‘data centre services market’, set to grow at 12.4% compound annual growth rate (CAGR) until 2022, hitting revenues of more than AU$2 billion by 2021.
By their very nature, modular data centres are packed with carefully engineered electrical and cooling components, meaning there are huge opportunities for professionals in these areas who want to avail themselves of the opportunities in this market.
Edging out costs
Aside from faster access to information, edge data centres help to keep operational costs low, making them more affordable options for customers.
These savings multiply as organisations invest in this infrastructure for more and more locations, as cooling costs, downtime and operational visibility through remote monitoring across the network increase. The designs have low power usage effectiveness (PUE) — sub 1.5 — to ensure minimal energy consumption and environmental impact. Designs also look at the ability to expand and contract within the structure, and also expand and contract as a structure.
Essentially, these devices give power back to customers to manage their own environment with little fuss. Everything from access control to power utilisation monitoring, temperature monitoring and inventory management is all controlled under the same umbrella.
At Vertiv, we ran a survey among IT, data centre and facilities professionals, which indicated cooling capacity, service and maintenance are top concerns. It also showed that the remote locations at which these data centres are most likely to be deployed tend to lack dedicated IT personnel, so central monitoring over regional data centres is incredibly valuable.
Keeping data centres cool
Getting cooling right helps reduce costs, speed and time to market, and mitigate downtime risks, and herein lies opportunities for those who can provide new approaches to cooling within edge data centres. The challenge on the edge is what technology to use and whether it meets business requirements (chilled water, direct expansion, evaporative cooling, etc). The ease of deployment and maintenance also come into consideration, particularly in non-metropolitan areas.
Direct expansion systems certainly allow for ease of deployment and also provide relatively simple and well-known installation and ongoing maintenance practices. One of the challenges in the past has been incorporating energy efficiency into these systems. However, the introduction of variable cooling capacity has helped enable this.
Going to the next level, modern design and newer architectures have enabled refrigerant economisation systems that further drive PUE efficiency with no involvement of water. This depends on the location and application of the edge data centre, but it can mean further energy savings of up to 50% with mechanical PUEs below 1.1, helping to both reduce costs and achieve efficiency goals. Another key area is being able to utilise lower outside ambient temperatures when available.
If the system can optimise operation for outdoor ambient temperatures as they change, IT loads can take advantage of 100% of the potential economisation hours. This ensures maximum efficiency — as needed, the system can turn off one or even both of its unit compressors, instead engaging the refrigerant pump, which operates at just 5% of the energy used by the compressors.
Built into these systems is a supervisory control system that contains no single point of failure and one single point of management for all cooling infrastructure. That means fast access to actionable data, system diagnostics and trending conditions.
The removal of points of failure, such as water piping systems, naturally improves reliability. If extra capacity is needed, edge data centres can add it through a scalable, modular design with no need for extra chillers, cooling towers or even ductwork. The need for regular manual adjustments involving water and outside air are also eliminated as everything is automated.
Monitoring in these systems can be done through simple mobile applications, which provide IT managers with remote visibility into thermal conditions inside small edge spaces. Some apps can even allow managers to track technician workflows and remotely control units when behind secure firewalls. This speeds up recovery while significantly reducing downtime. It also reduces the need for on-site service calls, the costs for which can increase quickly, and provides the opportunity for predictive maintenance.
Looking over the edge
Realistically, we’re still at the beginning of the phenomenon we call the edge, and what it means for how the data centre market will change in the future. As the proliferation of IoT devices continues, and the demand for data and speed increase, we may start to see more remote and even smaller edge computing spaces develop over time. Moore’s Law has yet to be proven wrong. Anticipating this is part one of the equation — realising and capitalising on the opportunity for electrical and cooling professionals will see it through.
OM5 multimode fibre (MMF) was introduced with fanfare in 2016, but to date has seen very few...
Asia Pacific is expected to have more than 8 billion connected devices by 2020. These devices...
The government's plan for the nbn is a mess and recent events provide an insight into why the...