dc-overview 1.0.5

GPU Datacenter Monitoring Suite - Prometheus, Grafana & Exporters

DC Overview

Complete GPU datacenter monitoring suite. Monitor your GPU servers with Prometheus, Grafana, and optional AI-powered insights.

✨ What's Included

Component	Description	Port
Prometheus	Time-series database for metrics	9090
Grafana	Beautiful dashboards and alerting	3000
node_exporter	CPU, RAM, disk, network metrics	9100
dcgm-exporter	NVIDIA GPU metrics (utilization, temp, power)	9400
dc-exporter	VRAM temperature, hotspot, fan speed	9500
vastai-exporter	Vast.ai earnings and reliability (optional)	8622

---

🚀 Quick Start

Prerequisites

• Linux (Ubuntu 20.04+, Debian, CentOS)
• Python 3.9+ with pip
• Root/sudo access for installing services

One Command Setup

Ubuntu 24.04+ / Python 3.12+ (uses pipx):

sudo apt install pipx -y
pipx install dc-overview
pipx ensurepath && source ~/.bashrc
sudo dc-overview quickstart

Ubuntu 22.04 / Python 3.10 (direct pip):

pip install dc-overview
sudo dc-overview quickstart

Alternative (if you get "externally-managed-environment" error):

pip install dc-overview --break-system-packages
sudo dc-overview quickstart

For remote worker deployment, set up passwordless sudo on workers:

sudo bash -c 'echo "YOUR_USER ALL=(ALL) NOPASSWD: ALL" > /etc/sudoers.d/nopasswd && chmod 440 /etc/sudoers.d/nopasswd'

The wizard guides you through everything:

╭──────────────────────────────────────────────────╮
│           DC Overview - Quick Setup              │
╰──────────────────────────────────────────────────╯

Step 1: What is this machine?
  ○ GPU Worker (has GPUs to monitor)
  ● Master Server (monitors other machines)
  ○ Both (has GPUs + monitors others)

Step 2: Setting up Monitoring Dashboard
  Set Grafana admin password: ******
  ✓ Prometheus running on port 9090
  ✓ Grafana running on port 3000

Step 3: Add Machines to Monitor
  How do you want to add servers?
    ● Import from file/paste (recommended)
    ○ Enter manually

  Paste your server list:
  global:root,mypassword
  192.168.1.101
  192.168.1.102
  192.168.1.103
  [Enter]

  Installing on 192.168.1.101... ✓
  Installing on 192.168.1.102... ✓
  Installing on 192.168.1.103... ✓
  ✓ Added 3 workers to Prometheus

Step 4: Vast.ai Integration (Optional)
  Are you a Vast.ai provider? [y/N]: y
  Vast.ai API Key: ******
  ✓ vastai-exporter running (port 8622)

✓ Setup Complete!
  Grafana: http://192.168.1.100:3000

---

📋 Import File Format

Create a simple text file to add many servers at once:

Option 1: Global credentials (same for all)

global:root,mypassword
192.168.1.101
192.168.1.102
192.168.1.103
192.168.1.104

Option 2: Per-server credentials

192.168.1.101,root,password1
192.168.1.102,ubuntu,password2
192.168.1.103,admin,password3

Option 3: Mixed (global default + overrides)

global:root,defaultpass
192.168.1.101
192.168.1.102,ubuntu,custompass
192.168.1.103

---

🔧 Manual Installation

On Master Server (monitoring hub)

pip install dc-overview
sudo dc-overview quickstart
# Select "Master Server"

On GPU Workers

pip install dc-overview
sudo dc-overview quickstart
# Select "GPU Worker"

Or from the master, provide SSH credentials and the wizard installs remotely.

---

📊 Available Commands

dc-overview quickstart          # ⚡ One-command setup (recommended)
dc-overview status              # Check what's running
dc-overview add-machine IP      # Add another machine to monitor
dc-overview install-exporters   # Install exporters on current machine
dc-overview setup-ssl           # Set up reverse proxy with SSL

---

🔒 Reverse Proxy & SSL Setup

Set up a secure HTTPS frontend with a branded landing page:

Self-Signed Certificate (Default)

# Basic setup (IP access only)
sudo dc-overview setup-ssl

# With custom site name
sudo dc-overview setup-ssl --site-name "My GPU Farm"

# Include IPMI Monitor
sudo dc-overview setup-ssl --ipmi --vastai

Let's Encrypt (Free SSL)

For a valid SSL certificate (no browser warnings):

sudo dc-overview setup-ssl \
  --domain monitor.example.com \
  --letsencrypt \
  --email [e-mail] \
  --ipmi --vastai

DNS Setup (Required for Domain)

Add these DNS records pointing to your server IP:

Type	Name	Value	Purpose
A	monitor.example.com	<server-ip>	Main dashboard
A	grafana.monitor.example.com	<server-ip>	Grafana subdomain (optional)
A	ipmi.monitor.example.com	<server-ip>	IPMI subdomain (optional)

After Setup

Access your monitoring at:

https://<server-ip>/           # Landing page
https://<server-ip>/grafana/   # Grafana dashboards
https://<server-ip>/prometheus/# Prometheus UI
https://<server-ip>/ipmi/      # IPMI Monitor (if enabled)

Or with domain:

https://monitor.example.com/
https://grafana.monitor.example.com/  (if subdomain configured)

---

🐳 Docker Alternative

If you prefer Docker Compose:

# docker-compose.yml
version: '3.8'
services:
  prometheus:
    image: prom/prometheus:latest
    ports: ["9090:9090"]
    volumes:
      - ./prometheus.yml:/etc/prometheus/prometheus.yml

  grafana:
    image: grafana/grafana:latest
    ports: ["3000:3000"]
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=admin

---

🔗 Related Tools

Tool	Purpose	Install
IPMI Monitor	Server health, SEL logs, ECC errors	pip install ipmi-monitor
dc-exporter	GPU VRAM temperatures	Included in quickstart

---

📖 Full Suite Setup (Master + Workers)

For a complete datacenter setup with IPMI monitoring:

1. On Master Server

# Install dc-overview (Grafana + Prometheus)
pip install dc-overview
sudo dc-overview quickstart
# Select "Master Server", add your workers

# Install ipmi-monitor (optional - for BMC/IPMI)
pip install ipmi-monitor
sudo ipmi-monitor quickstart

2. Workers are configured automatically

The quickstart installs exporters on workers via SSH.

3. Import your servers

Create servers.txt:

global:root,sshpassword
192.168.1.101
192.168.1.102
192.168.1.103

Then paste when prompted, or run:

dc-overview add-machine 192.168.1.101 --ssh-pass mypassword

---

💬 Support

• Discord: https://discord.gg/7yeHdf5BuC
• Issues: https://github.com/cryptolabsza/dc-overview/issues

---

📄 License

MIT License - see LICENSE for details.