Data center capacity will ramp up in 2021

Europe’s four leading data center markets (London, Amsterdam, Frankfurt, Paris) are expected to see 415 megawatts of capacity coming online in 2021, a 22.5% increase over the previous year’s total supply. This year of record growth is forecasted to be followed by two more of similar magnitude, with 370 and 440 megawatts coming online in 2022 and 2023, respectively (all figures CBRE Data Centre Solutions, Q4 2020).

These remarkable figures are hardly coincidental. As we emerge from the global pandemic, enterprises are poised for acceleration in digital transformations and the related adoption of cloud technologies – driving high demand for data center resources. That is, after all, where compute and storage resources for applications, data, and content are located. In the years to come – in the new, cloud-based, application-centric world emerging from the COVID-19 related disruption – access to that data and that processing power will become even more of a business-critical challenge than it is already today.

All these dynamics make high-capacity enterprise WAN networks essential to business survival – and underscore the importance of their reliability, scale and deterministic performance. Corporate data centers need to talk to each other – to share data and content, and provide back-ups for redundancy. High-speed connectivity services enable the transit of those assets over any distance – from across town to ultra-long trans-oceanic links, and everything in-between. With a fast, reliable connection in place, physically separate corporate data centers can easily share resources and balance workloads.

An additional, and increasingly significant aspect playing into this context is sustainability. While data centers have a positive environmental role to play – through the digital ecosystems they facilitate – they are also major contributors to the global carbon footprint, reportedly emitting as much CO2 as the airline industry. As sustainability climbs up corporate agendas, enterprises will increasingly move their data to data center facilities powered by renewable energy sources, which may be further away from company locations. This further raises the importance of the corporate WAN, and especially how predictably high-capacity links perform. Our recent study suggests that the sustainability angle is already a major are of consideration, with 95% of enterprises saying they would pay more to secure a sustainable supplier of network services.

A robust enterprise WAN and connectivity strategy must ensure a safe, quick route for your most important data. In this context, optical services have emerged as a mission-critical resource, simply too important to fail. When building your end-to-end WAN strategy, three themes come to the fore as the most important aspects from the high-performance perspective:

CAPACITY AND SCALABILITTY

Data sets coming into or leaving a data center can be very large—from tens of Gigabytes to Terabytes and beyond. At the top end of the enterprise market, routers sending that data are gradually moving from 100GbE to higher speed 400GbE services. At the lower end, we see companies moving from multiple 10GbE links to 100GbE. To handle these growing bandwidth demands, your network strategy needs to include reliable, high-capacity connections that are able to scale simply and quickly, allowing you to address not just today’s, but also tomorrow’s needs.

LATENCY PARAMETERS

Minimizing latency is crucial to the successful implementation of data center applications such as database mirroring. Excessive latency can be a particularly dangerous phenomenon in an enterprise WAN, as it creates bottlenecks that can delay, or even prevent, data from reaching its destination. When bottlenecks occur, additional packets often need to be sent, potentially creating congestion on the network and affecting application performance.

Latency from the physical medium, the optical fiber, is – naturally – critical and needs to be taken into consideration in network design. But one additional reason those crucial, high-capacity network links typically stay at the optical layer of the OSI stack (outside of the sheer volume of data they often transport) is that optical services are inherently deterministic – the physical path between end points is known, the channel is not shared with any other traffic, and therefore the associated latency can be reliably predicted. They also eliminate another, related potential problem – jitter, or latency variability.

MAXIMUM UPTIME AND CUSTOMER SUPPORT

Much as equipment manufacturers and telecom service providers dislike admitting it, things sometimes do go wrong, even in core networking technologies engineered to five or six “nines” availability (99.999% availability, or “five nines”, assumes just 5 minutes 15 seconds of downtime a year). The freak wintry spell that Texas experienced at the beginning of 2021 is a stark reminder of that – due to power blackouts critical network connections for many enterprises went down, causing business disruption on top of a humanitarian crisis.

Maximum uptime is critical to business continuity – half of enterprises we recently surveyed say the impact of corporate WAN outage exceeding 24 hours would be catastrophic. That is why planning is necessary for more than just the “steady state” when considering your enterprise WAN. Protected, route-diverse services are the first line of defense when it comes to unexpected faults. Equally importantly, however, it is worth taking an active interest in how your service provider’s support organization performs – and especially, how quickly problems typically get solved.

When it comes to their most critical network links, enterprises deserve the best. Optical services provide exactly that – they occupy the core of the cloud-centric world, supplying network capacity for users and applications to connect to storage, compute, and content resources that are the lifeblood of business. Telia Carrier and Ciena have joined forces to help you be on top of your game – visit our landing page to learn more and see how we can help future-proof your most critical network resource.

Bernd Hoogkamp, Head of Products