Enterprise-grade containerization
Instead of spending a week setting up a Kubernetes cluster — run it in a minute. Our VPS images and dedicated servers ship with a production-ready container environment with monitoring, networking and storage configured. We've been running it since 2019 — thousands of Polish-company apps run in production on this setup.
What you get from minute one
Docker 24+ ready
Docker Engine, Docker Compose, BuildKit, local registry for your company's images. Logging to Loki, metrics to Prometheus configured out of the box.
k3s (lightweight K8s)
Full Kubernetes implementation tuned for VPS. Single-node or multi-node cluster. Runs on 2 GB RAM — pick a larger plan for heavy workloads.
kubectl + helm
CLI tools, helm charts repository, Cert-Manager for auto-SSL on the cluster. helm 3 and kustomize supported.
Container registry
Private registry (Harbor) for your images. Push from GitHub Actions, GitLab CI, Jenkins. CVE image scanning built in.
Ingress + Traefik
Reverse proxy ready to go, with automatic Let's Encrypt certificate issuance. Wildcard SSL, HTTP/3 and custom middleware supported.
Prometheus + Grafana
Container monitoring out-of-the-box. CPU, memory, network and disk metrics. Loki for logs. Pre-configured alerts to Slack/email.
Quick start — your first app in 4 steps
After SSH-ing into a fresh VPS, here's what you need to run your first containerized app. Every command works out of the box — no install required.
Check k3s is running
kubectl get nodesYou should see one node in Ready status. If you see an error, k3s is starting in the background — wait 30 seconds.
Deploy an app via Helm
helm install my-app oci://harbor.4hosting.net/charts/wordpress --version 18.0.0Pulls the official WordPress chart, deploys deployment + service + ingress + PVC. Live in ~90 seconds.
Configure DNS and SSL
kubectl apply -f - <<EOF apiVersion: cert-manager.io/v1 kind: Certificate ...EOFCert-Manager automatically issues Let's Encrypt SSL for your domain. Just point an A-record to your VPS IP.
Scale the app
kubectl scale deployment my-app --replicas=5Five replicas of the same app behind a load balancer. Traefik distributes traffic automatically.
Bare VPS · k3s on VPS · Managed K8s — which to pick?
Each option makes sense at a different scale. Small projects often run better on a single VPS with Docker. For medium and large — k3s on VPS or a dedicated server. Managed Kubernetes (EKS, GKE) makes sense only at 100+ pods.
| Feature | VPS + Docker | VPS + k3s | Managed K8s |
|---|---|---|---|
| Price (monthly) | from 59 PLN | from 119 PLN | from 400 PLN + traffic |
| Setup | 5 minutes | 5 minutes (preinstalled) | 15-30 minutes |
| Auto-scaling | Manual | HPA + VPA | HPA + VPA + Cluster |
| Multi-tenancy | Hard | Namespace + NetPol | Full |
| For whom | Single apps | 5-50 microservices | 100+ pods, multi-region |
Common use cases
Microservices in production
App split into 5-50 services, each in a container, with auto-scaling and blue-green deployment. Service mesh via Cilium or Linkerd. Internal traffic encrypted with mTLS.
CI/CD pipeline
GitHub Actions → Docker build → push to registry → kubectl rollout. No developer VPS in the loop. One-click rollback to the previous version.
Multi-tenant hosting
Agency managing 50+ clients — each in an isolated namespace with its own RAM/CPU limits. NetworkPolicy isolates traffic between clients.
Dev / staging / prod
Three clusters on one dedicated server. Identical configuration, isolated environments. Promote staging to prod with one `kubectl apply`.
Migrating to our stack
We help migrate containerized apps from other platforms for free. Whether you use Docker Compose, Heroku, Render, Vercel or Cloudways — we have migration experience and know where the gotchas hide.
Full DevOps stack ready to go
Frequently asked questions
Is k3s production-ready?
Yes. k3s is the official, certified Kubernetes distribution from Rancher (part of SUSE). Used in production by thousands of companies. It differs from full Kubernetes only by being lighter — the entire K8s API is available.
What if I need more nodes?
k3s scales to multi-node easily. Just order a second VPS and join it to the cluster with one command. Longhorn distributed storage works across nodes automatically.
Does support help with setup?
Yes, on managed plans (VPS 32G and up). We configure ingress, SSL, monitoring and storage for your app. On smaller plans we provide ready Helm charts and docs.
Can I use Docker Swarm instead of K8s?
You can, but we don't recommend it. Docker Swarm has been practically deprecated since 2020 — Mirantis announced EOL. k3s has the same simple API as Swarm but is actively developed.
How is Kubernetes resource billing handled?
We bill for VPS resources (RAM, vCPU, NVMe), not pods or containers. You can run 1 or 100 pods on one VPS — it doesn't change the price.
Spin up your cluster in 60 seconds
Pick a VPS NVMe from 59 PLN/mo. SSH in and run your first `kubectl apply`. Free migration from other platforms — we help move manifests and data.
See VPS plans →
