Build vs. Buy - Evaluating Edge Solutions

September 16, 2020

At Section, we are increasingly being approached by enterprises whose bespoke edge needs have them asking themselves whether to build vs. buy when it comes to edge compute capabilities.

These enterprises are weighing up:

  • If they can achieve the customization they need to suit their unique requirements;
  • The elapsed time and product and engineering effort it will take to build out Edge due to its complex and diversified nature;
  • The diverse infrastructure relationships they will need to maintain;
  • The skills they will need to hire and maintain to build and operate the networking, compute and orchestration technologies; and
  • Whether it makes sense to pursue a build strategy in an area which is most often outside the core business.

Types of Edge Use Cases

In a recent report, MarketsandMarkets projected the entire edge computing market will grow from $3.6B in 2020 to $15.7B by 2025 at a CAGR of 34.1%.

Drivers of this include the growing adoption of IoT across industries, increasing demand for low-latency processing and real-time automated decision-making solutions, and an urgent requirement for surmounting hugely increasing data volumes and network traffic.

As edge computing is increasingly prioritized as a key strategic initiative, the types of enterprises we are seeing approach Section regarding their custom edge needs include:

  • Software providers who are losing business because they don’t have a cloud offering and need to SaaS-ify their product offering (e.g. WAF providers with only on-prem solutions).
  • Hosting providers looking to expand their footprint and capabilities with edge app hosting.
  • PaaS providers looking to build or augment their solutions with edge services.
  • Enterprise engineering teams who don’t have, or don’t want to invest in, the resources required to build and operate advanced edge provisioning, orchestration, scaling, routing, and monitoring.

Edge Computing “Build” Pioneers

Some of the pioneers of edge computing, such as eBay, Netflix and Chick-fil-A have made serious investments in edge engineering teams and technology to meet their data, latency and availability needs. Let’s look briefly at the investments in edge technologies that each of these edge pioneers have made.

eBay

Back in 2018, eBay announced a three year plan to replatform and modernize its backend infrastructure, including decentralizing its cluster of data centers and moving to an edge computing architecture. eBay’s primary servers were originally located in Phoenix, the Salt Lake City environs, and across Nevada. The company is now in the midst of moving these online services and data closer to users distributed across the US. The goal: “to create a faster, more consistent user experience, saving 600-800 milliseconds of load time.” eBay simultaneously announced a build of its own custom-designed servers. Overhauling the physical eBay infrastructure also necessitated a redesign and rebuild of the entire logic and software stack that runs on the infrastructure. No small task. This is a long-term move that has required significant investment from the company.

Netflix

Netflix has spent years building infrastructure for content delivery on a global scale, including a significant core deployment on AWS and a CDN using thousands of Open Connect caching servers distributed across data centers and ISPs in over 200 sites around the world. The streaming giant is still in the midst of its edge computing transformation. Recently, its Vice President of Networks, Dave Temkin, said the streaming giant was looking at edge computing for its production side of the house to alter how content is shared during series production. In TV and film production, edge computing can perform “data thinning”, i.e. transcoding, to compress large datasets down to smaller files, which can be sent across the network for review by creatives on the other side of the world.

Chick-fil-A

Back in 2018, Chick-fil-A announced it was in the middle of an extensive edge computing strategy, setting out to run a Docker-based edge compute infrastructure on 6,000 devices in all 2,000 of its restaurants around the US. The problem they were facing was the fact that many of their restaurants were doing over three times the volume of foot traffic than they were originally designed for. One of the ways Chick-fil-A set out to solve this was by “invest[ing] in a smarter restaurant.” The goal: to improve customer service and operational efficiency through a data collection and analytics strategy rooted in “making smarter kitchen equipment” and developing an IoT platform for Chick-fil-A’s restaurants.

As these examples highlight, some of the arguments for building your own edge infrastructure might include:

  • Greater predictability;
  • More control and flexibility; and
  • Potential development and operational efficiencies.

Build vs. Buy

However, some would make the argument, in the case of a business like Chick-fil-A, why would a fast food restaurant choose to become a technology company? The time and investment involved in building and running your own edge infrastructure, including orchestrating resources, managing scaling, routing, and monitoring, is immense. Replatforming is not for the faint-hearted and many SMBs and large enterprises have still not migrated to the cloud, let alone been able to access the latency and availability benefits of edge computing.

At the time of the Chick-fil-A edge deployment, there weren’t many (if any) options to shop the required solutions out, so they were forced to build. There is also often the misconception that building automatically gives you the best solution, and buying means you’re not innovative. Now, however, edge platforms like Section make it possible to build on top of proven edge solutions so companies that want to innovate can do so while focusing on their core business, thus accelerating their path to the Edge.

Section’s Adaptive Edge Engine (patent-pending) uses machine learning to intelligently optimize edge workload placement based on real-time traffic demands. Our Adaptive Edge Engine abstracts the underpinning decision-making, execution and monitoring to offer developers a reliable, turnkey solution, which optimizes application performance. This will lead to more efficiencies, which will have an impact not just on budgets, but also promote greater sustainability.

Working with Section to Build Your Edge Application, SaaS or PaaS

Internet users are demanding more secure and faster digital experiences. However, SaaS, PaaS and application providers are frustrated with the amount of time and technical capability required to move their applications closer to their end users, which can deliver those better digital experiences.

Section is an Edge Compute platform which enables any provider to quickly build their own Edge Solution without needing to worry about the massive complexities (and resources involved) of distributed provisioning, orchestration, scaling, monitoring and routing. With Section you can deploy an Edge application in minutes or an Edge PaaS or SaaS in hours. You will also gain the 24x7x365 support of our dedicated team of engineers.

Sign up online or talk to a solutions engineer to get started with Section.