The Nirvana Examples library is live

April Wong

May 21, 20262 min read

TL;DR: 18 ready-to-clone Terraform templates for running databases, AI, Kubernetes, dev infra, observability, networking, and web workloads on Nirvana. Browse, fork, terraform apply. → github.com/nirvana-labs-examples

Today we're opening the door on something the team's been building quietly for a while: a library of production-shaped examples for everything from Postgres to LangChain. Each one is a public GitHub template — fork it, change the inputs, run it.

What's in it

A scan of the catalog:

Databases — Postgres, MongoDB (single-node and a 3-node replica set), Redis (cluster, plus a version on NKS), ClickHouse
AI / RAG — a self-hosted LangChain + LangGraph stack on a single Nirvana VM, plus a customer-support agent built on the same pieces
Kubernetes — ArgoCD on NKS for GitOps deployments, MongoDB on NKS, Redis on NKS
Dev infrastructure — GitHub Actions self-hosted runners, GitLab CE, HashiCorp Vault
Observability + messaging — Prometheus + Grafana, RabbitMQ
Networking — Tailscale subnet router, WireGuard VPN server
Web — Caddy with automatic HTTPS, Node.js + PM2 + Nginx

Browse them all: github.com/nirvana-labs-examples

Why a library, and not just more docs

Documentation tells you what's possible. Templates show you what works.

Every example here is a real deployment pattern — the configuration we'd use if we were standing it up ourselves. Terraform handles the infrastructure; Ansible or Helm handles the config depending on the example. Everything is templated so you can run it as-is or fork it for your own setup.

The cloud-specific bits — Nirvana VMs, ABS volumes, networking — live inside each template. The application layer doesn't change. If you've stood any of these tools up on EKS or GKE before, the manifests you wrote there largely work here too.

How to use it

Each repo is a GitHub template — one click gives you your own copy. From there:

bash

git clone <your-fork>
cd <your-fork>
terraform init
terraform apply

Specifics for each example (variables, prerequisites, what to expect) live in the repo's README.

What's next

This is part of a longer campaign. Up next: a deep dive on the marquee NKS example — an end-to-end application running on the cluster, persistent volumes on ABS, ingress, the works. After that, category-by-category spotlights over the coming weeks.

Star the org to follow along: → github.com/nirvana-labs-examples

If you try something and hit a wall, open an issue. Feedback now shapes what we build next.