Q&A: What is Edge-as-a-Service (EaaS)?
As technology advances, consumers are demanding faster, more personalized, and more secure digital experiences, and it’s a steep challenge for existing centralized and cloud application delivery models to keep up with these expectations.
In order to solve this puzzle, many companies are moving services to the edge. However, the majority of organizations simply don’t have the resources or the expertise to build or manage the complex distributed systems required for effective edge delivery.
For these companies, edge-as-a-service (EaaS) provides a one-stop solution to accelerate their path to edge. Innovative edge technologies offload the complexities associated with moving applications to the edge, including infrastructure provisioning, workload orchestration, scaling, monitoring and traffic routing. With EaaS, companies of all sizes can quickly and cost-effectively access the edge to deliver the digital experiences customers expect.
In this Q&A, Stewart McGrath, CEO of Section, discusses edge technology and how EaaS can provide solutions for companies seeking help moving to the edge.
Q: How can a company benefit from EaaS? What’s the business value?
A: The business value of edge has been well documented. We know that providing customers with faster, richer digital experiences, improving security perimeters and saving data backhaul can all deliver better business outcomes.
We have also seen point solution software providers leverage edge as a new deployment paradigm for their software. As a natural deploy location for web software solutions for security, API gateways, performance optimization and more, edge technology allows these software providers to not only let their customers consume their software as a service without paying a ‘latency penalty’ but can also then couple their solutions with complementary edge services to provide a one-stop security and optimization solution for their customers.
Q: Knowing the benefits of Edge, the next question must be: What is the benefit from an “as-a-service” model for edge?
A: There are a number of advantages to as-a-service technologies, and they apply to the process of moving services to the Edge alongside the process of consuming software-as-a-service. IBM outlined 5 Key Benefits of Software as a Service, which all stem from the ability to leverage technology provided and supported by a third party. The speed to deploy, cost advantages from reduced Day 2 operations and reduced scalability or update challenges all relate to the ability of an EaaS provider to deliver a pre-baked best-of-breed solution versus the alternative of building it yourself.
What IBM does not mention, but which I believe is also a critical advantage to as-a-service solutions, is the ability to focus on core business rather than being distracted by all the elements necessary to deliver services from the edge. Delivering distributed systems from the edge is highly complex and unlikely to be subsumed as a core business capability except within the very largest of enterprises. Being able to consume complex but non-core services as-a-service can help organizations to stay hyperfocused on their core business.
Q: What kinds of companies use EaaS?
A: We have seen both software and infrastructure providers leveraging EaaS solutions to present their offerings as edge-enabled.
As noted above, software providers are continuing to focus on building out their core software capabilities and leveraging EaaS so that they can improve the consumption model for their software as well as easily couple additional services to their core offerings.
Infrastructure providers are using EaaS to both present their footprint as a cohesive, general-purpose workload Edge solution and also to augment their existing footprint with additional locations without needing to deploy hardware in those locations.
Q: What part does Kubernetes play at the edge for EaaS?
A: Kubernetes can play an outstanding role as an orchestrator of workload within an edge cluster. For a massively multi-tenanted EaaS cluster, the control, security and flexibility Kubernetes provides is critical.
Orchestration of workloads at edge locations can be achieved with Kubernetes, but we need to consider abstractions from underlying infrastructure when considering deployments of an edge onto federated compute or various providers. We also need to account for overall traffic management and total edge load balancing in order to deliver full EaaS.
When considering running an edge which is laid over the top of a range of disparate infrastructure providers, we have to look at creating an environment whereby the vagaries of the underlying infrastructure are removed so that we can create a common Kubernetes operating model on all target locations. This means not consuming specific Kubernetes services from specific providers (such as those provided by AWS, Google and others). If EaaS solutions were to rely on one specific provider for Kubernetes solutions, those Edge solutions would either be locked to specific providers or have significant operational challenges with different Day 2 operating environments for every provider.
While Kubernetes creates a wonderful scaling and orchestration base within a cluster, an EaaS solution should also include an overarching scaling and routing strategy to manage the entire network size, location and traffic routing. To use an analogy, if we think of the complete edge network as a single computer, Kubernetes is controlling each CPU (edge cluster) and an overall operating system is required to orchestrate, manage and optimize workloads across that distributed computer.
Q: Can other CNCF projects play a part in EaaS?
A: A number of graduated and incubating CNCF projects are lining up beside Kubernetes to solve various requirements of EaaS solutions.
For example, one of the challenges presented by distributed systems is the collation and storage of metrics and logs from the distributed nodes. Customers of EaaS should expect to be able to consume quality, near-time, centralized telemetry and diagnostics from that EaaS solution. Projects such as Prometheus for telemetry and FluentD for diagnostics fit very well into solving for this category.
Service mesh and tracing projects as well as CoreDNS and proxy solutions such as Envoy also fit extremely well in the technology stack of an edge-as-a-service offering.
Edge-as-a-service can be helpful in providing a one-stop solution to accelerate a company’s path to edge technology. Both software providers and infrastructure providers are currently considering the value that EaaS can bring. By offloading many of the complexities associated with moving applications to the edge including infrastructure provisioning, workload orchestration, scaling, monitoring and traffic routing, companies of all sizes can quickly and cost-effectively access the edge to deliver the digital experiences customers expect by leveraging EaaS. CNCF projects are playing a critical role in delivering edge-as-a-service and we expect them to continue to do so well into the future.
To hear more about cloud-native topics, join the Cloud Native Computing Foundation and the cloud-native community at KubeCon+CloudNativeCon North America 2021 – October 11-15, 2021