Applying Kubernetes in Edge and IoT applications – Challenges and Solutions
Kubernetes describes itself as "a portable, extensible open-source platform for managing containerized workloads and services".
It was originally developed to manage workloads running in large clouds – but has an extensible architecture which makes Kubernetes a potential general purpose foundation for building a declarative API based control plane for “things” beyond just containerized workloads.
The established momentum of Kubernetes, along with a the growing “landscape” of collateral tools (logging, tracing, service mesh, more) are causing many to evaluate whether Kubernetes is adaptable as a control plane for managing and operating workloads and devices at edge locations.
Operating at edge has many similarities with cloud computing, but also has challenges including security, messaging, resource constraints, data communications, application lifecycle management, and more.
This session will survey of some existing solutions in the edge and IoT space, mapped to three approaches that are commonly considered:
- Run Kubernetes full clusters at edge locations (e.g. retail store and remote branch office) under tight resource constraints
- Run containerized workloads at edge with a remote Kubernetes control plane at a cloud or central location
- Run devices using non containerized software at edge