AI Chat Startup Guide (FastAPI + React/Vite + Caddy)

This project is a lightweight chat client designed to run local AI models via Ollama. The entire stack is containerized using Docker and runs on top of lightweight Alpine Linux images.

---

🚀 Quick Start

Prerequisites

1. Docker and Docker Compose installed on your host machine.
2. Ollama installed locally on your PC

---

Step 1. Configure Local Ollama on Host (for Linux/Fedora)

By default, Ollama on Linux only listens on 127.0.0.1 (localhost). To allow the backend Docker container to communicate with it, configure Ollama to accept connections from the Docker network:

1. Create a service override configuration file:

   sudo mkdir -p /etc/systemd/system/ollama.service.d && echo -e "[Service]\nEnvironment=\"OLLAMA_HOST=0.0.0.0\"" | sudo tee /etc/systemd/system/ollama.service.d/override.conf

2. Apply the configuration changes and restart the Ollama service:

   sudo systemctl daemon-reload && sudo systemctl restart ollama

3. Verify that Ollama is listening on all network interfaces:

   ss -tlnp | grep 11434

   You should see *:11434 or 0.0.0.0:11434 instead of 127.0.0.1:11434.

4. Allow port 11434 in your Firewall (if UFW / firewalld is active): If UFW is active, incoming connections from Docker containers to port 11434 will be blocked by default. Allow the port:

   sudo ufw allow 11434/tcp

---

Step 2. Set-up your data base

1. change file name from .env-test to .env

2.in file put your data

test_DB_HOST=
test_DB_PORT=
test_DB_USER=
test_DB_PASS=
test_DB_NAME=

SECRET_KEY =
ALGORITHM = HS256
ACCESS_TOKEN_EXPIRE_MINUTES = 3200

---

Step 3. Start the Stack via Docker Compose

Run the build and start all services in the background:

docker compose up -d

Upon initial startup, Docker will automatically:

• Pull the required Alpine-based images.
• Build the React frontend in a temporary container and copy the static assets to Caddy.
• Verify PostgreSQL and Redis readiness via built-in healthchecks.
• Run database migrations using Alembic.
• Spin up Prometheus, Grafana, and Loki with pre-configured settings.

---

🔗 Available Services

Once successfully started, the services will be accessible at the following URLs:

Service	Browser URL	Description
AI Chat (Frontend)	🌐 http://localhost	Caddy serves the React application on the standard port 80.
API Documentation	📖 http://localhost/docs	Interactive Swagger UI for testing API endpoints.
Grafana	📊 http://localhost:3000	Metrics dashboard. Automatically logs in as Admin.
Prometheus	⏱️ http://localhost:9090	Time-series database containing metrics.
adminer	💻 http://localhost:8080/	admin panel in browser.

---

📊 Pre-configured Grafana Monitoring

Monitoring dashboards are set up and imported automatically:

1. Open Grafana (http://localhost:3000).
2. Navigate to Dashboards in the left sidebar.
3. Open the Application folder and select AIWEBHOST Chat Dashboard.
4. You will see real-time charts displaying:
   • User activity and message count.
   • Response latency from Ollama models.
   • Redis cache hits/misses and PostgreSQL database metrics.

---

📊 Admin-Panel settings

convenient admin panel in your browser

1. Open Adminer (http://localhost:8080).
2. Login using:
   • System: PostgreSQL
   • Server: db
   • User: postgres
   • Password: 1
   • Database: ai-slop

---

🛠️ Useful Management Commands

• Check container status:

  docker compose ps

• Stream backend logs in real-time:

  docker compose logs -f backend

• Restart a specific service (e.g., Caddy):

  docker compose restart caddy

• Rebuild the project after code changes:

  docker compose up --build -d

• Stop and remove all containers:

  docker compose down

---

⚙️ Ports Configuration Notes (For Developers)

To prevent conflicts with services that might already be running on your local machine:

• Redis inside Docker does not expose port 6379 to the host. The backend container communicates with it directly inside the internal Docker network.
• PostgreSQL is forwarded to port 5433 on the host machine (instead of the standard 5432). If you need to connect to the database directly from PyCharm, pgAdmin, or another client, use port 5433 (username: postgres, password: 1, database: aiwebhost).

---

👥 Contributions

• Backend (FastAPI): Built by me (exception: send_message endpoint there AI helped with images).
• Frontend (Vite + React): Fully vibecoded
• DevOps (Docker, Caddy, Grafana, Prometheus): AI helped with configuration.

AI Tools & Skills Used:

• Gemini
• vercel-labs / find-skills
• mattpocock / grill-me