Bringing Kubernetes to the Edge is Becoming Table-Stakes for Equipment Vendors

The operational proposition for deploying at the edge is clear, especially for data intensive verticals and those looking towards Industry 4.0 and enterprise-grade 5G: it reduces backhauling, improves end-user performance, and enables low latency use cases. But its highly distributed architecture will always create control and orchestration problems. So, how should edge nodes be efficiently deployed and controlled? ABI Research believes that Kubernetes must play a role. Equipment vendors (IT vendors and telecom infrastructure providers) certainly think so too.

In Mid-2021, Intel rebranded their edge orchestration portfolio, and announced Intel Smart Edge Open (royalty free) and Intel Smart Edge (optimized for commercial application). These tools are part of a service aimed at helping end-users run and manage applications/network functions at the edge of the network from a central control panel. Nokia offers a similar service that leverages Open Application Program Interfaces (APIs) and Kubernetes to provide a cloud-native network automation function that aids efficient deployment at the edge. Ericsson’s Cloud Container Distribution utilizes similar tools to manage and orchestrate all containerized applications deployed in its edge products.

A critical component of Intel’s Smart Edge management tool is a Kubernetes-powered Cloud Native Stack, but what is Kubernetes? Kubernetes is an open-source platform for managing containerized workloads and services. It orchestrates distributed workloads and provides a single management point, while automating manual processes involved in deploying, managing, and scaling edge nodes within the hybrid cloud. ABI Research believes that Kubernetes will bring enormous benefits to the edge for CSPs and enterprises:

• Orchestration Centralization—Reduces asymmetric deployment of applications and protocols within edge nodes; converges workloads, reduces network complexity; automation of previously manual tasks.
• Scalability—Cloud-nativity ensures that infrastructure can dynamically scale to control containers regardless of growth.
• Enable Use Cases—Advanced 5G use cases, e.g., AI/ML require low latency and network consistency. Kubernetes will manage data containers, ensure that network performance is maintained, and facilitate service deployment in the edge.
• Assist AI/ML Workloads—Kubernetes management tools will ensure that these workloads are simply and optimally orchestrated to maximize value extraction.
• Drive Go-to-Market Time—Kubernetes will reduce go-to-market as it enables efficient and automated provisioning of applications/instructions to edge nodes.
• Maturity—Kubernetes can deliver a trusted software that ensures dynamic control at the edge.

IT Vendors

A Kubernetes-powered control/management software will be a critical enabler of service deployment at the edge within small and medium enterprises (SMEs). Without specialized or dedicated network management teams, SMEs struggle to implement a highly distributed network of edge nodes. But, by utilizing Kubernetes, they could be able to easily deploy and control edge nodes from central platforms. This will have significant commercial value, as it will optimize networking costs and performance, while also enabling SMEs to integrate latency intensive AI/ML technologies, which have the potential to revolutionize business operations. ABI Research would go as far as to suggest that without Kubernetes-powered visibility and control in edge nodes, for SMEs, Industry 4.0 technologies will remain elusive, and they will not maximize operational value. This should send a signal to IT vendors highlighting the need to embrace Kubernetes as integral to the design of edge infrastructure and encourage enterprise adoption.

Moreover, Kubernetes could be a gamechanger for enterprise control and configuration of private 4G and 5G networks. It enables an open, cloud-native, telecom-grade architecture that reduces complexity and backhauling. These synergies offer IT vendors the opportunity to introduce commercially valuable ecosystems that enable the deployment ‘controlled’ private networks at the edge.

Telecom Infrastructure Providers

For telecoms, Kubernetes are fast becoming a de facto standard and it is critical for infrastructure providers to embrace this trend. On the one hand, it will allow them to drive value-add services at the edge in conjunction with their 5G networks. This will allow telecoms support end-to-end deployments and provide an ecosystem-first approach to value-add services. These services will be critical revenue drivers for telecoms, especially given the ongoing decline in the cost of connectivity. On the other hand, Kubernetes’ control of edge node deployments will help 5G meet required use cases, as it can help telecoms provide the promised latency, availability, and consistency required for vertical specific cases. As telecoms recognize the opportunities offered by Kubernetes, it is critical for infrastructure providers to integrate it, drive usability/functionality, and provide end-to-end network visibility.

Driving Kubernetes to the edge is necessary, but equipment vendor will continue to face significant challenges:

1. Deployment—Kubernetes requires highly skilled managers and a significant time commitment to integrate it into the edge. Offering this as part of an infrastructure driven service can be costly, especially with on-going skills shortages,
2. Cost—Deploying Kubernetes across edge locations will be expensive, and these costs are likely to ‘skyrocket’ as edge locations multiply. For equipment vendors, this will likely lead to increased maintenance and service-architecture costs. With traditional Capital Expenditure (CapEx), passing these costs onto the enterprise can often discourage investment, especially given the on-going market standardization of Kubernetes.

ABI Research strongly believes that Kubernetes will unlock value at the edge, and to drive its value we recommend that equipment vendors evaluate the following commercial strategies:

• Deploy Kubernetes ‘as-a-service’. This will drive enterprise deployment (preference for Operating Expenditure (OpEx)), while enabling equipment vendors to pass on on-going costs to the enterprise.
• Develop vendor-agnostic, open Kubernetes edge controller that can be deployed universally independent of equipment.
• Bundle private network with Kubernetes-powered edge deployments. Driving ecosystems and ‘value-add’ services will be critical revenue growth generators for equipment players.
• Shift ‘go-to- market’ to target use cases rather than proving the technology to telecoms and enterprises.