Price Bot

Production-ready Telegram bot with modular architecture, optional gRPC database microservice, observability stack, tests, and CI/CD.

• Docs: Architecture, gRPC API, Observability

Features

• Aiogram 3.x router/action/schema separation
• Pydantic v2 settings and validation
• orjson for fast JSON serialization
• Optional gRPC DB microservice (SQLAlchemy 2.x)
• Dockerized dev and prod setups
• Observability: Loki, Promtail, Prometheus, Grafana
• Health and metrics endpoints: /healthz, /ready, /metrics
• Tests (pytest) and Ruff linting
• GitHub Actions CI/CD

Architecture

• bot/src: bot app (routers, actions, schemas, misc, services)
• database: gRPC DB microservice (models, repository, server)
• monitoring: configs for Loki/Promtail/Prometheus/Grafana
• docker-compose.*.yml: base, local overlay, prod overlay

Requirements

• Docker and Docker Compose
• Python 3.11 for local dev/tests

Setup

cp .env.example .env

Update .env with your secrets (TOKEN, Postgres, etc.).

Run (local)

docker compose -f docker-compose.yaml -f docker-compose.local.yml up -d --build

• Bot: polls Telegram, exposes /healthz, /ready, /metrics on 8080
• Database service: gRPC on 50051
• Postgres: 5432

Run (prod)

docker compose -f docker-compose.yaml -f docker-compose.prod.yml up -d --build

Reads configuration from .env.

Monitoring

docker compose -f docker-compose.yaml -f docker-compose.local.yml up -d loki promtail prometheus grafana

• Grafana: http://localhost:3000 (admin/admin)
• Prometheus: http://localhost:9090
• Loki readiness: http://localhost:3100/ready See docs/observability.md for details.

Testing

python -m pip install -r requirements.txt
pytest -q

Postgres-backed tests expect TEST_DATABASE_DSN or running local Postgres (see CI).

Linting

ruff format .
ruff check .

CI/CD

• .github/workflows/ci.yml: Ruff + pytest + Docker build
• .github/workflows/cd.yml: Build & push images to GHCR on main/master/prod

License

MIT