TSI-Telemetry Backend Setup on Proxmox LXC

This document outlines the setup for the TSI-Telemetry project's backend components within a Proxmox LXC container, leveraging Docker for TimescaleDB and a Python script for MQTT data ingestion. This setup was implemented to run the telemetry pipeline on a dedicated server environment, decoupling it from the Mac Mini control plane.

1. lxc-docker Container Setup

Given the inability to enable KVM virtualization, a dedicated LXC container (lxc-docker) was created and specially configured to host Docker and its containers.

Container Details:

• Hostname: lxc-docker
• IP Address: 192.168.68.30/24
• Resources: 2 Cores, 2GB RAM, 32GB Disk
• Key Feature: Created with nesting and keyctl features enabled in Proxmox to allow Docker to run inside.

Docker Installation: Inside the lxc-docker console, Docker was installed using the official convenience script:

apt update && apt upgrade -y
curl -fsSL https://get.docker.com -o get-docker.sh
sh get-docker.sh

2. TimescaleDB Deployment

The TimescaleDB instance, crucial for time-series data storage, was deployed as a Docker container within lxc-docker.

Deployment Command:

docker run -d --name timescaledb \
  -p 5433:5432 \
  -e POSTGRES_PASSWORD="YOUR_STRONG_PASSWORD" \
  -v timescaledb_data:/var/lib/postgresql/data \
  --restart unless-stopped \
  timescale/timescaledb:latest-pg15

Database Initialization: After deployment, the telemetry database and timescaledb extension were created:

docker exec -it timescaledb psql -U postgres -d postgres
# In psql:
CREATE DATABASE telemetry;
\q

docker exec -it timescaledb psql -U postgres -d telemetry
# In psql:
CREATE EXTENSION IF NOT EXISTS timescaledb;

CREATE TABLE car_metrics (
  time           TIMESTAMPTZ NOT NULL,
  rpm            INTEGER,
  speed          INTEGER,
  coolant        INTEGER,
  intake_temp    INTEGER,
  throttle       INTEGER,
  engine_load    INTEGER,
  map_kpa        INTEGER,
  fuel_level     INTEGER,
  timing_adv     INTEGER,
  battery_voltage DECIMAL(4,2)
);

SELECT create_hypertable('car_metrics', 'time');
\q

3. Python MQTT Bridge Deployment

The Python script (bridge.py) acts as the intermediary, subscribing to HiveMQ Cloud and inserting data into TimescaleDB.

Installation & Configuration (lxc-docker console):

# Install Python and dependencies
apt update
apt install python3 python3-pip -y
pip3 install paho-mqtt psycopg2-binary

# Create script directory
mkdir -p /opt/telemetry-bridge

# Create bridge.py (content copied from bridge-script.md)
# Ensure HiveMQ and TimescaleDB credentials are updated within the script.
nano /opt/telemetry-bridge/bridge.py

systemd Service Setup (lxc-docker console): The script was configured to run as a robust systemd service:

cat << EOF > /etc/systemd/system/telemetry-bridge.service
[Unit]
Description=TSI Telemetry Bridge
After=docker.service
Requires=docker.service
Wants=network-online.target
After=network-online.target

[Service]
ExecStart=/usr/bin/python3 /opt/telemetry-bridge/bridge.py
Restart=always
User=root
StandardOutput=journal
StandardError=journal
RestartSec=5

[Install]
WantedBy=multi-user.target
EOF

systemctl daemon-reload
systemctl enable --now telemetry-bridge
systemctl status telemetry-bridge

4. Grafana (lxc-monitoring) Integration

A separate container (lxc-monitoring) was set up to host Grafana for data visualization.

Container Details:

• Hostname: lxc-monitoring
• IP Address: 192.168.68.22/24

Grafana Installation: Installed using the official Grafana APT repository.

Data Source Configuration: A PostgreSQL data source was added in Grafana, connecting to 192.168.68.30:5433 (the lxc-docker container), database telemetry, user postgres, and the TimescaleDB password. The "TimescaleDB" toggle was enabled.

Dashboard Import: A pre-generated JSON model was imported to instantly create the full "TSI Telemetry" dashboard, featuring gauges for real-time metrics and a time-series graph.