Mind the gap – how close is your data to the edge?

The buzz around technologies like 5G, edge computing, and IoT is insatiable at the moment, with talks of how it can expedite a move towards smart cities, improved and more personalised customer experiences, and businesses utilising advanced and artificial intelligence. Much of this change can be attributed to the rise of cloud computing. But business leaders are often neglecting one of the most important questions – how am I ensuring the speed and quality of my data transfers between where it is gathered, processed, analysed, and stored? This aspect of any digital transformation initiative is key to a successful implementation that optimises your investments in these new technologies.

Understanding your first mile

Let’s look at edge computing as an example – a term that can actually be quite misleading. When you say ‘edge’, you imagine a network that extends to the furthest boundaries of a coverage area, when in actual fact an edge networking device can be as close as the checkout counter of your supermarket.

The first mile of connectivity is the distance from an actual device – think point of sale (PoS) machines – to the actual ‘edge’ of an organisation’s network and its secure systems, usually housed in a data centre. Data normally travels this first mile over public internet infrastructure, instead of a private, secure network connection. This not only opens up gaps in security, but it also brings up quality-of-service concerns because routing over the internet can be unpredictable.

Most organisations are already using technologies such as SD-WAN to extend the edge of their private networks closer to these “edge” devices so that less data has to traverse a best-effort internet connection. However, this often comes with large costs and added complexity and can still leave data to cross vast suboptimal distances to get from the point of collection to its final destination.

Why edge computing is an imperative for innovation in a data-driven world

Claudio Scola, head of product management EMEA at Lumen Technologies, discusses the importance of edge computing for innovation in a data-driven world. Read here

Securing the middle- and last-mile

One of the biggest issues with allowing data to travel a large distance over the public internet is security. Simply put, the longer data is out there over a vulnerable path, the longer it can be a target. It becomes vulnerable to DDoS attacks like route hijacking, which can have serious business implications.

Every organisation should be getting their apps and data onto private infrastructure as quickly as possible. That means actually bringing your network’s edge as close as possible to where data is being created or collected, minimising the time spent traversing the public internet and maximising the security.

SD-WAN provided the first step in this journey, creating reassurance for mission-critical data as organisations opened branches and expanded their footprint to better serve distributed workforces and applications during the COVID-19 pandemic. It also continues to be a big priority for organisations even as the workers begin to return to offices after vaccinations with 82% of decision-makers saying SD-WAN would be a top organisational priority in 2021 in a recent Forrester survey.

A lesser talked about technology might be the key to unlocking the potential of technologies like SD-WAN, 5G, edge computing, and IoT. Network Functions Virtualisation (NFV) hosting pushes the limits of modern networks by allowing companies to create an on-demand virtual edge to their networks, closing that gap between an enterprise’s network and the various edge devices where transactions and/or data from mission-critical applications are created. Having NFV as part of your SD-WAN and cloud connectivity strategy enables traffic to make the smallest numbers of hops on the public internet before it passes through a private, secure software-defined network (SDN). In other words, you shorten that first mile, and optimise the middle- and last-mile by having your data travel safely and quickly over a private connection.

Think about every time you tap your card at a PoS terminal on a shopping trip or a night out. To process this transaction, your financial details need to travel to a central point and back again. How long do you want those details traveling across public infrastructure? With a virtual edge, this distance can be significantly minimised, providing predictable, secure connectivity.

Neglecting connectivity should give you the jitters

The other downside of using the Internet is its unpredictable routing, which can cause jitter-sensitive applications to underperform.

Jitters cause variances in the transmission of transaction data – meaning that sales could potentially be delayed. Across many retail facilities and thousands of customers a day, this can soon add up to significant lost annual revenue from transactional latency.

However, by deploying a virtual edge that is highly distributed for localised traffic, you are essentially bringing the edge closer to where the transaction takes place. This creates predictability within the network, minimising jitter and reducing latency.

How the edge and the cloud tackle latency, security and bandwidth issues

Mark Seddon, CEO of Pact Global, discusses how edge computing is complementing the cloud to tackle latency, security and bandwidth issues. Read here

Time to close the gap

You can deploy all the trendy new technology that’s out there, and you still might not get the economic benefits you seek. You can plug in your POS system, connect it to 5G, link it up to your SD-WAN edge device so you can track and orchestrate data on the edge of your network. But if the way you’re getting all that data back to your network is insecure and inefficient, you’ll still have a suboptimal solution.

NFV hosting technologies are emerging to help improve edge connectivity, closing the gap between your network’s edge and the data you process. These technologies can be deployed on demand, in a point-and-click manner, like spinning up a cloud instance. It takes minutes, not months, to improve network performance and matches the agility of your business, and you won’t need a project plan to roll it out. This not only makes it more efficient, but also reduces costs by drastically reducing the need for hardware and avoiding the duplication of overhead and resources.

As you deploy new technology, think about the gap it is creating between the true edge of your infrastructure and where the data is being collected. Every mile of connectivity must be accounted for, so don’t forget to mind the gap.

Written by Paul McGuinness, head of solutions EMEA at Megaport

Editor's Choice

Editor's Choice consists of the best articles written by third parties and selected by our editors. You can contact us at timothy.adler at stubbenedge.com