Private preview opening soon

Run Kubernetes across edge, cloud, and bare metal.

Otroid provisions immutable k0s clusters on your own infrastructure using Kairos images, P2P coordination, FluxCD GitOps, Cilium networking, and Kubernetes-native lifecycle operations.

Notify Me at Launch See the Flow

AWS GCP Azure Hetzner Bare Metal ARM64 + AMD64

otroid.cluster.yaml

Declared production stack

ready

runtime: k0s
os: kairos
image:
  channel: nightly-tested
  variant: hardened
  arch: [amd64, arm64]
network:
  discovery: p2p
  cni: cilium
gitops:
  engine: fluxcd
  repo: customer/platform-state
day2:
  upgrades: kairos-operator
  backups: etcd-snapshots
isolation:
  virtualClusters: vcluster
  hardenedNodes: vnode

BOOT

cloud-config injected

JOIN

peers coordinate

SYNC

Git reconciles

PRIVATE PREVIEW WAITLIST

Be first in line when Otroid opens.

We are inviting teams who need Kubernetes outside the managed-cloud happy path: edge sites, multi-cloud VPCs, ARM fleets, bare metal, and constrained environments.

USER JOURNEY

From cloud access to declared state.

Otroid keeps the path intentionally narrow: connect infrastructure, import the right images, boot immutable nodes, let peers coordinate, and hand the cluster to GitOps.

01_CONNECT

Attach provider

Select AWS, GCP, Azure, Hetzner, or bare metal. Otroid provides scoped IAM policy YAML and verifies access.

02_IMPORT

Prepare images

Machine images are imported for quick future cluster creation across cloud regions and architectures.

03_BOOT

Inject config

First boot receives cloud-config, credentials, coordination settings, and the selected platform manifests.

04_FORM

Coordinate peers

Nodes discover each other over P2P links using coordination servers in your VPC or hosted by Otroid.

05_SYNC

Reconcile stack

FluxCD installs Cilium, Hybrid networking, Automation credentials, metrics, ingress, and any chosen manifests.

CORE PLATFORM

Opinionated where production usually hurts.

Otroid is not a marketplace for every possible Kubernetes add-on. It ships the common production path, keeps the stack declarative, and leaves the cluster open for normal Kubernetes extension.

Immutable Kairos Images

Kairos-init and Dockerfile-built OCI images produce repeatable nodes with base and hardened variants.

Nightly Integration Testing

Images are built continuously and validated before becoming candidates for new clusters or upgrades.

k0s + Cilium

A focused Kubernetes runtime and eBPF network layer, with no CNI sprawl in the first release.

FluxCD by Design

Create a new Git repository or connect an existing one, then let cluster state reconcile from Git.

vCluster + vNode Isolation

Create virtual clusters for teams and workloads, with stronger node isolation in hardened images.

Secure Boot Ready

Designed for tamper-resistant node lifecycle, hardware-backed trust, and safer edge deployments.

Multi-Cloud and Bare Metal

Use the same platform model across VPCs, regions, clouds, edge sites, and physical machines.

CAPI or k0rdent Lifecycle

Cluster lifecycle management is treated as platform plumbing, not a manual runbook.

DAY-2 OPERATIONS

Lifecycle controls without server drift.

Once the cluster exists, Otroid keeps upgrades, snapshots, restores, and rebuilds tied to declared state instead of hand-maintained machines.

> UPGRADE

Kairos Operator fetches the target image tag and restarts nodes immediately or on schedule.

> BACKUP

Capture etcd snapshots and restore cluster state when recovery matters.

> DESTROY

Tear down cleanly without leaving hidden coordination or cluster resources behind.

> EXTEND

Keep using standard Kubernetes primitives, manifests, CRDs, and Git workflows.

upgrade-plan.yaml scheduled

cluster: edge-prod
strategy:
  mode: rolling
  window: "02:00-04:00"
nodes:
  selector:
    role: worker
targetImage:
  repo: ghcr.io/otroid/node
  tag: kairos-k0s-hardened-2026.06.12
preflight:
  - etcd-snapshot
  - cilium-health
  - flux-reconcile
rollback:
  keepPreviousImage: true

YOUR INFRASTRUCTURE

Run across your cloud, edge, and on-prem environments.

Bring the environments you already operate: public cloud accounts, private networks, edge sites, and bare-metal estates. Otroid connects to each location with scoped access, keeps workloads in your infrastructure, and supports hybrid topologies when clusters span networks or providers.

> ACCESS

Connect cloud accounts, private networks, and bare-metal environments using scoped, auditable access instead of broad platform credentials.

> VALIDATE

Verify provider permissions, image availability, network reachability, and required automation hooks before a cluster is created.

> CONNECT

Support hybrid topologies across VPCs, clouds, edge sites, and on-prem networks with customer-owned or Otroid-hosted coordination.

> INTEGRATE

Wire DNS, certificates, Git providers, ingress, and automation credentials only when the selected cluster profile needs them.

LAUNCHING SOON

A narrow, production-ready path to Kubernetes anywhere.

Join the preview list for immutable clusters across cloud, edge, and bare metal.

Join the Waitlist