Eventer

Telegram-first private event management platform.

Structure

Path	Purpose
Backend/	NestJS monorepo — apps/api, apps/bot, apps/worker + shared libs
Frontend/	Next.js Organizer/Admin dashboard
*.md	Product/tech blueprint (start with 00-README.md)

Local setup

# Infra
docker compose up -d postgres redis

# Backend
cd Backend
cp .env.example .env   # adjust DATABASE_URL / secrets
pnpm install
pnpm prisma migrate deploy
pnpm start:api:dev     # http://localhost:3000/health  +  /docs
# optional: pnpm start:bot:dev / pnpm start:worker:dev

# Frontend
cd Frontend
cp .env.example .env
pnpm install
pnpm codegen           # types from Backend/openapi/openapi.json
pnpm dev               # http://localhost:3001

Full app stack via Compose: docker compose --profile full up --build

Payments (OrcaRail)

Eventer supports PAYMENT_PROVIDER=mock (default) or PAYMENT_PROVIDER=orcarail.

1. Copy Backend/.env.example → .env and set:
   • PAYMENT_PROVIDER=orcarail
   • ORCARAIL_API_KEY / ORCARAIL_API_SECRET
   • ORCARAIL_TOKEN_ID / ORCARAIL_NETWORK_ID (catalog UUIDs from OrcaRail)
   • ORCARAIL_RETURN_URL / ORCARAIL_CANCEL_URL (e.g. http://localhost:3001/payments/return and .../cancel)
   • ORCARAIL_WEBHOOK_SECRET (dashboard webhook secret, or your API secret)
2. In the OrcaRail dashboard, point the webhook URL to
   https://<your-eventer-api>/payments/webhook/orcarail
   (signature header: x-webhook-signature).
3. Self-hosted OrcaRail: set ORCARAIL_BASE_URL=https://your-orcarail-host/api/v1 — no code changes required.

Checkout flow: create intent → confirm → open hosted pay_url (Telegram bot sends a Pay button; Frontend has return/cancel pages). Ticket issuance stays webhook-driven (payment_intent.completed).

Production checklist

Derived from 12-deployment-observability.md:

1. Secrets — load JWT_SECRET, TELEGRAM_BOT_TOKEN, R2_*, payment keys, SENTRY_DSN from a secrets manager (never bake into images).
2. Migrate separately — run pnpm prisma migrate deploy (or Backend/scripts/migrate-and-start.sh only for single-replica staging) before rolling new api/bot/worker replicas.
3. Processes — deploy four independently scalable processes: api, bot, worker, web (see Dockerfiles under Backend/ and Frontend/).
4. Health — GET /health (liveness), GET /health/ready (DB readiness), GET /metrics (basic Prometheus text).
5. Bot mode — long polling in non-production when TELEGRAM_BOT_TOKEN is set; production should use webhook mode (token still required).
6. Backups — daily Postgres backups with ≥7-day retention; test restore before go-live. Prioritize audit_logs and payments.
7. Observability — structured JSON logs; optional SENTRY_DSN (stub logs until @sentry/node is wired); alert on queue age, 5xx rate, DB pool exhaustion.
8. CI/CD — Backend: lint → test → build → migrate deploy → roll images. Frontend: lint → codegen → test → build → deploy. Staging auto on main; production via manual promotion.