Skip to content

DEPLOY.md

Latest commit

 

History

History
98 lines (72 loc) · 4.16 KB

DEPLOY.md

File metadata and controls

98 lines (72 loc) · 4.16 KB
Raw

Deploying the hosted report

Two deployables:

  • API (FastAPI engine) → a container host. This guide uses Fly.io (Dockerfile + fly.toml are ready); Railway works from the same Dockerfile.
  • Frontend (web/, Next.js) → Vercel.

Optional but recommended for production:

  • Upstash Redis — shared cache + per-IP throttle across instances.
  • A GitHub token (PAT or GitHub App installation token) so the API runs on the 5 000 req/hr authenticated limit instead of 60 req/hr.

1. API → Fly.io

# One-time
fly launch --no-deploy            # or `fly apps create ghra-report-api`
fly volumes create ghra_data --region iad --size 1   # persists the waitlist DB

# Secrets (never commit these; set them on Fly)
fly secrets set GHRA_GITHUB_TOKEN=ghp_xxx
fly secrets set GHRA_CORS_ORIGINS=https://your-frontend.vercel.app
# If using Upstash (see §3):
fly secrets set GHRA_REDIS_URL=rediss://default:xxx@xxx.upstash.io:6379

fly deploy

fly.toml already wires the health check (GET /api/health), the /data volume mount, and the non-secret config (GHRA_REPORT_TTL_SECONDS, GHRA_RATE_LIMIT, GHRA_RATE_WINDOW_SECONDS, GHRA_WAITLIST_DB=/data/waitlist.db).

The container runs uvicorn with --forwarded-allow-ips=*, so behind Fly's proxy the per-IP throttle keys on the real client address (no GHRA_TRUST_FORWARDED_FOR needed).

Verify: curl https://ghra-report-api.fly.dev/api/health → {"status":"ok","github_token":true}.

2. Frontend → Vercel

cd web
vercel link
vercel env add NEXT_PUBLIC_API_BASE production   # → https://ghra-report-api.fly.dev
vercel --prod

Set the project Root Directory to web/ in Vercel (the repo root is the Python engine). After the frontend URL is known, set it as GHRA_CORS_ORIGINS on the API (§1) so the browser's cross-origin calls are allowed.

Vercel commit-author gotcha: if vercel --prod is blocked on the commit author, deploy from a git-free copy of web/ and vercel alias set.

3. Upstash Redis (production cache + throttle)

Without GHRA_REDIS_URL the API uses an in-process store — correct, but per-instance (cache and throttle don't share across machines). For more than one instance, create an Upstash Redis database and set its rediss:// URL as GHRA_REDIS_URL (§1). The hosting extra (redis) is already installed in the image. Any Redis server version works (the throttle uses plain EXPIRE).

4. Environment reference

Variable Where Default Purpose
GHRA_GITHUB_TOKEN API (none) Server token → 5 000 req/hr + GraphQL repo lists.
GHRA_CORS_ORIGINS API localhost:3000 Comma-separated allowed browser origins.
GHRA_REDIS_URL API (in-memory) Upstash/Redis URL for shared cache + throttle.
GHRA_REPORT_TTL_SECONDS API 3600 Report cache TTL.
GHRA_RATE_LIMIT API 20 Requests per window per IP.
GHRA_RATE_WINDOW_SECONDS API 3600 Throttle window.
GHRA_WAITLIST_DB API <output>/waitlist.db SQLite path (point at the mounted volume).
GHRA_TRUST_FORWARDED_FOR API off Only if not using uvicorn --forwarded-allow-ips.
NEXT_PUBLIC_API_BASE Frontend localhost:8080 API base URL the browser calls.

5. Notes & follow-ups

  • Waitlist durability: SQLite on the mounted Fly volume survives restarts. For multi-instance writes, migrate the waitlist to Postgres (Neon) — only SqliteWaitlistStore needs a sibling implementation behind the existing WaitlistStore protocol.
  • Local parity: run the API with uv run --extra serve python -m uvicorn --factory src.serve.app:create_app --port 8080 and the frontend with pnpm dev in web/.