To say Kubernetes[1], everyone's top container orchestration pick, is hard to master is an understatement. Kubernetes doesn't have so much as a learning curve as it does a learning cliff. But, Canonical[2]'s MicroK8s[3] lets you learn to climb it in your home. And, with its latest release, it's easier than ever to set up a baby Kubernetes cluster using inexpensive Raspberry Pi[4] or NVIDIA Jetson[5] single-board computers (SBC).
MicroK8s is a tiny Kubernetes cluster platform. You can install it on Ubuntu Linux[6], or other Linux distros, macOS, and Windows. With it, you can get your feet wet with Kubernetes on your own desktop and office.
Also: Best Raspberry Pi alternatives: Banana Pi, Odroid, NanoPi, and more[7]
MicroK8s is more than just a Kubernetes training arena. It's powerful enough with its strict confinement to be used in Internet of Things (IoT)[8] and edge computing[9] applications. In this context, strict confinement means your MicroK8s cluster has complete isolation from the underlying operating system. With strict confinement and MicroK8s add-ons, such as Istio[10], Knative[11], CoreDNS[12], Prometheus[13], and Jaeger[14], this means you can use MicroK8s to make small footprint Kubernetes clusters suitable for edge gateways. Making it even more useful, you can deploy Kubernetes clusters securely at the edge on x86 and ARM hardware with a single Snap[15] command.
To make it even handier both as a training environment and as a platform for deploying IoT and edge applications, MicroK8s 1.21 memory footprint has been shrunk by 32.5%[16], as benchmarked against single node and multi-node