An immutable Linux distribution that boots into a Kubernetes node.
A complete single-node cluster, in full:
#cloud-config
corium:
role: singleThat is the entire interface for the common case. Every other field has a default, and every default can be overridden.
Read the docs · Quick start · Source
The operating system is a container image. Corium is built with a
Containerfile, pushed to a registry, and versioned by digest. You inspect it
with podman, scan it with the scanner you already run, promote it between
environments by moving a tag, and roll it back atomically. There is no separate
image-building toolchain to learn, because the one you use for your applications
already works.
Kubernetes ships with the OS. k0s lives in the
read-only /usr. Upgrading Kubernetes means booting a new OS image — one
version axis, one upgrade mechanism, one rollback path. A node is a disposable
artefact rebuilt from a digest, not a machine that accumulates state.
Configuration is cloud-init. Not a bespoke API, not a new configuration language: the mechanism every cloud and hypervisor already speaks. Where cloud-init does not exist — bare metal, PXE, a preconfigured appliance — the same schema is read from a file, the kernel command line, or a default baked into the image. Every field is documented in the configuration reference.
Corium is not a fleet-management control plane; it provisions nodes and stops there. It does not fork or patch k0s. It does not invent a configuration language. Those omissions are the point: the project is an opinionated integration, and its value is in what it declines to do.
Talos Linux is the dominant option here and it is excellent — it is also an API-driven system with no shell and its own configuration model, which is a real commitment. Kairos covers similar ground, but deliberately builds its own A/B partition scheme on an arbitrary base distribution rather than using the OSTree/bootc lineage.
Corium's bet is narrower: that for teams already living in OCI registries and GitOps, an operating system that is an image — built, signed, scanned and promoted like every other image they ship — is worth more than a bespoke mechanism, however good.
Early. The architecture is settled and every claim above has been exercised on real hardware: single nodes, unattended ISO installs, and a three-controller highly available control plane whose virtual IP was verified by hard-stopping the controller holding it. It is not ready for anything you would miss.
| Documentation | Quick start, configuration reference, and the decisions behind the project |
| Corium-OS/Corium | The operating system: Containerfile, agent, and deployment scripts |
| Examples | Single node, workers, add-ons, a custom CNI, and a highly available control plane |
| Architecture decisions | Why fedora-bootc over Fedora CoreOS, and why ext4 over xfs |
Licensed under MIT.
