np — Nano Platform
Deploy like Kubernetes. Simple like Compose.
np is a deployment platform for small-to-medium teams that need production-grade capabilities without production-grade complexity.
Quick Start
# 1. Install
curl -fsSL https://np.run/install.sh | bash
# 2. Initialize
np init
# 3. Deploy your first service
np deploy my-app
# 4. Check status
np status
# 5. View logs
np logs my-appWhy np?
| If you use… | You get… | But miss… |
|---|---|---|
| Docker Compose | Simplicity | Health checks, auto-restart, multi-node |
| Kubernetes | Full orchestration | Complexity, resource overhead, ops burden |
| Bare Nomad | Scheduling | Battery-included experience |
| np | Single binary. Real platform. | Nothing you need. |
What’s in the box
np bundles the best open-source infrastructure into one coherent platform:
| Layer | Component | Role |
|---|---|---|
| Scheduler | Nomad | Job placement, rolling updates, health checks |
| Service Discovery | Consul | Service registry, health checking, DNS |
| Ingress | Traefik | Reverse proxy, automatic TLS, routing |
| Observability | OpenObserve | Logs, metrics, dashboards — one pane of glass |
| CLI | np |
Single command to rule them all |
Managed
np status — know exactly what’s running where.
[OK] api-gateway 3/3 running, healthy
[OK] user-service 2/2 running, healthy
[OK] auth-service 1/1 running, healthy
[OK] frontend 1/1 running, healthyLightweight
Runs on a Raspberry Pi. Or a 10-year-old server. Or your laptop.
- Single binary CLI (< 20 MB)
- No dependencies beyond Linux
- 100 nodes, 500+ services without tuning
Batteries Included
No assembly required. No YAML engineering. No glue code.
# np.yaml — your entire deployment config
name: my-app
binary: ./my-app
health_check: /health
resources:
cpu: 100m
memory: 128mbOne file. One command. Done.
When NOT to use np
np is purpose-built. It’s not for everything.
- ❌ Not for 1000+ node clusters — that’s Kubernetes territory
- ❌ Not for multi-cloud orchestration — use Terraform + K8s
- ❌ Not for serverless/FaaS — different paradigm
For teams with 3-15 servers running 20-200 services, np is the right tool.