## The short version

Deploy a public directory built with Rails on Ample using the relational records pattern. Compute runs the app in an isolated microVM behind a public HTTPS URL, a managed PostgreSQL 16 database is auto-provisioned and injected as DATABASE_URL, and a published bucket serves public assets from the CDN host.

## WHAT WAS VERIFIED

Verified on Rails: a workspace-scoped table, an explicit state machine (draft to review to published) that rejects invalid transitions, and a cross-workspace read that returns nothing (transitions=true isolation=true).

# INFRASTRUCTURE

## What it needs

- [Compute](/content/discover/primitive/compute/index.html): verified (Apps run in isolated x86_64 Firecracker microVMs; sizes are the priced VM sizes.)
- [Postgres](/content/discover/primitive/postgres/index.html): verified (Managed PostgreSQL 16 runs in its own microVM and is auto-provisioned when an app needs a database and no DATABASE_URL is supplied.)
- [CDN](/content/discover/primitive/cdn/index.html): verified (A published bucket's objects are served at its public_url. It does not front app compute and is not a cache or key-value store.)

## BEFORE YOU START

- A Rails project that builds and starts with the documented commands (bundle install into vendor/bundle from Gemfile.lock (development and test groups skipped), then Puma via `rails server` reading PORT on the ruby-3.4 template with RAILS_ENV=production)
- A PostgreSQL driver reading DATABASE_URL at runtime (auto-provisioned when omitted, or supplied with --env)
- A bucket from `ample bucket create` with its credentials passed as encrypted S3_* environment variables
- An Ample account token with servers:write, databases:read, buckets:read

## How to do it

1. **Build and start**  
   bundle install into vendor/bundle from Gemfile.lock (development and test groups skipped), then Puma via `rails server` reading PORT on the ruby-3.4 template with RAILS_ENV=production. The server must bind 0.0.0.0 on PORT.

2. **Implement the pattern on PostgreSQL**  
   The fixture's module implements relational records: a workspace-scoped table, an explicit state machine (draft to review to published) that rejects invalid transitions, and a cross-workspace read that returns nothing (transitions=true isolation=true). Copy the approach into your schema; keep migrations idempotent and run them with --release-command.

3. **Wire object storage**  
   Create the bucket(s), then pass endpoint, region, bucket and keys as --env values. Use path-style addressing. Keep private data in an unpublished bucket. Publish only the bucket that serves public assets and reference its CDN URL.

4. **Deploy**  
   Run the synchronous deploy once and read the result (exit 0 live, 1 failed, 2 blocked). Re-running with no change is a no-op.
   
   ```  
   ample deploy . --name <app-name> --public --env S3_ENDPOINT=... --env S3_REGION=... --env S3_BUCKET=... --env S3_ACCESS_KEY_ID=... --env S3_SECRET_ACCESS_KEY=... --env CDN_PUBLIC_URL=...  
   ```

5. **Verify**  
   Fetch the live URL and the pattern self-test route(s) (/p/relational-records) from the example; then run your own checks. On failure read `ample logs <deployment_id> --kind build` then `--kind runtime`.
   
   ```  
   ample logs <deployment_id> --kind build  
   ```

## Tested examples

- **Rails pattern fixture** (`tests/deploy-canaries/rails-patterns`): Multi-pattern Rails app whose relational records module was checked live; the module is under tests/deploy-canaries/_pattern-modules.

## How to know it worked

- app responds on its public URL (`/` on the live URL, expect ample canary rails patterns)
- relational-records self-test (`/p/relational-records` on the live URL, expect records=ok transitions=true isolation=true)

## Cost estimate

Estimated 10.00 USD per month (size prices from pricing.toml at build revision ca0b484a7a9957c573a88d3a32dc603c3c10135a).

- app server x1 `s-1vcpu-1gb`: 5.00 USD
- managed PostgreSQL database x1 `s-1vcpu-1gb`: 5.00 USD

Upper-bound monthly price of the tested sizes. Buckets are allocation-priced per quota and not included.

## Verification evidence

- canary_run on 2026-09-20T21:55:28Z at revision `4f7cb9813c1f (CLI 4f7cb98)`: Rails pattern fixture deployed on Ample (bundle install into vendor/bundle from Gemfile.lock (development and test groups skipped), then Puma via `rails server` reading PORT on the ruby-3.4 template with RAILS_ENV=production); checks passed for relational-records and configuration-secrets. Pattern proof: a workspace-scoped table, an explicit state machine (draft to review to published) that rejects invalid transitions, and a cross-workspace read that returns nothing (transitions=true isolation=true). (expires 2027-03-19T21:55:28Z)
- canary_run on 2026-09-20T21:55:28Z at revision `4f7cb9813c1f (CLI 4f7cb98)`: The same Rails app deployed with a --release-command migration; the marker it created was readable after activation. (expires 2027-03-19T21:55:28Z)

## MCP tool

MCP tool `ample_deploy` (registry `mcp:ample_deploy`), schema hash `876465fce906da0c` observed 2026-09-22T00:31:22.774679+00:00 at revision `e6446ceea69b`. Binding state at export: schema_changed. Required scopes: servers:write, databases:read, buckets:read.