Guides on Corium

Concepts

Mon, 01 Jan 0001 00:00:00 +0000

The model underneath everything else. Read this once and the quick start and reference stop being surprising.

Quick start

Mon, 01 Jan 0001 00:00:00 +0000

From nothing to a working Kubernetes node in four steps, and a fifth for managing it without SSH.

Upgrades

Mon, 01 Jan 0001 00:00:00 +0000

Kubernetes ships with the operating system, so upgrading either means booting a new image. One version axis, one mechanism, one rollback path.

Cilium

Mon, 01 Jan 0001 00:00:00 +0000

Corium ships kube-router because k0s does: one component covering networking, network policy and service proxying. Replacing it with Cilium takes two decisions, and both fit in the cloud-config the node already boots with.

Software RAID

Mon, 01 Jan 0001 00:00:00 +0000

Corium can build software RAID arrays from a node’s spare disks, declared in the same corium: block as everything else. Putting the root filesystem on RAID is a different problem with a different answer, and this page covers both.

Building your own image

Mon, 01 Jan 0001 00:00:00 +0000

Corium’s published image is a base. Deriving your own is how you add the things a corium: block was never going to cover — a monitoring agent, your organisation’s CA bundle, a driver, a backup client. It is three lines of Containerfile for the simple cases.

WireGuard overlay

Mon, 01 Jan 0001 00:00:00 +0000

Corium has no native field for a host WireGuard interface yet. The design is decided but not yet built (ADR 6) — and for overlays that are not the cluster’s own transport it stays optional even once it ships. Until the field lands you build the interface the way you build anything the corium: block does not cover: with the escape hatches every node always has, write_files and runcmd. This page is that recipe.