# Host rest api with Express: Postgres-backed state

Deploy an Express REST API whose durable state lives in a managed PostgreSQL 16 database. With no DATABASE_URL supplied, Ample provisions the database and injects the connection string as an encrypted environment variable; the API stores rows with pg and serves them from a public HTTPS URL.

**Verified** November 6, 2026

## What it runs on

- [Computeverified](/content/discover/primitive/compute/index.html)  
Apps run in isolated x86_64 Firecracker microVMs; sizes are the priced VM sizes.

- [Postgresverified](/content/discover/primitive/postgres/index.html)  
Managed PostgreSQL 16 with pgvector runs in its own microVM and is auto-provisioned when an app needs a database and no DATABASE_URL is supplied.

## You need

- A package.json with a start script that listens on process.env.PORT (default 3000) and 0.0.0.0
- The pg driver (or another PostgreSQL client) reading process.env.DATABASE_URL
- Tables created by the app or by a --release-command migration, not assumed to exist
- An Ample account token with servers:write and databases:read

## Exact tested configuration

```bash
build npm run build --if-present
install npm install
runtime node
sizes-1vcpu-1gb
start npm run start
template node-22
```

## Steps

1. **Check the start script.** Ample runs npm install, npm run build --if-present and npm run start on the node-22 template. The server must bind 0.0.0.0 and read PORT.

2. **Read DATABASE_URL at runtime.** Create a pg Pool from process.env.DATABASE_URL. Do not hard-code localhost credentials; a local .env is replaced by the managed database in production.

3. **Deploy.** Run the existing synchronous deploy once from the project directory and read the result. Exit 0 is live, 1 is failed, 2 needs an answer first (a preflight question). Re-running with no change is a content-addressed no-op.

```bash
   $ ample deploy . --name <app-name> --public --format json
   ```

4. **Verify.** Fetch the live URL and run the success checks below. On failure read the build log, then the runtime log, fix the cause and deploy again; do not blind-retry.

```bash
   $ ample logs <deployment_id> --kind build
   ```

## How it was checked

✓ root route responds · http get/ → “ample canary express postgres api”  
✓ database insert and count succeed · http get/items → “postgres=ok”

## Evidence

1. November 6, 2026 · canary run  
   Express REST API using pg with no DATABASE_URL supplied: Ample auto-provisioned a managed PostgreSQL 16 database, injected the connection string, and /items reported postgres=ok after a real insert and count.

2. November 6, 2026 · canary run  
   Node app needing PostgreSQL with no DATABASE_URL: a managed database was auto-provisioned and /db reported postgres=ok.

3. November 6, 2026 · canary run  
   Express app with npm install and npm run start deployed on the node-22 template; the public URL served the expected response.

4. September 20, 2026 · live deployment  
   Authorized end-to-end run through the catalog: public search, verified contract, side-effect-free plan (ready_for_approval with a signed plan token), operator approval, handoff to ample_deploy, live URL, and the recipe's success checks passed.

## Tested examples

- **Express + pg canary.** Express API with POST /items and GET /items backed by an auto-provisioned managed PostgreSQL database.

## What it doesn’t cover

- Verified on the node-22 template at s-1vcpu-1gb.
- Managed PostgreSQL 16 with pgvector.
- Apps and their managed databases are placed together, in the same region.
