Watchtower

Made by Camila "Mocha" Rose

A simple, easy to use system resource monitoring solution

Perfect for servers • Powerful for everyone

---

Installation & Setup

Prerequisites

For C++ Collector:

# Ubuntu/Debian
sudo apt-get update
sudo apt-get install g++ libcurl4-openssl-dev libjsoncpp-dev

# Arch
sudo pacman -Syu
sudo pacman -S gcc curl jsoncpp

# macOS (requires Homebrew)
brew install curl jsoncpp

For Python Dashboard:

# Python 3.8 or higher required
pip install flask requests plotly

Running the System

Step 1: Start the Go Aggregation Server

cd backend/go
go run server.go

Step 2: Start the Python Dashboard

cd backend/python
python dashboard.py

Step 3: Start the C++ Collector

cd cpp-collector
g++ -std=c++17 -o metrics_collector metrics_collector.cpp -lcurl -ljsoncpp -lpthread
./metrics_collector http://localhost:8080/metrics 5

Configuration

Alert Thresholds (Python Dashboard)

Edit dashboard.py to modify alert thresholds:

ALERT_THRESHOLDS = {
    'cpu': 80.0,      # CPU usage threshold (%)
    'memory': 85.0,   # Memory usage threshold (%)
    'disk': 90.0      # Disk usage threshold (%)
}

Data Retention (Go Server)

Modify retention period in server.go:

maxAge: 24 * time.Hour,  // Keep metrics for 24 hours

Collection Interval (C++ Collector)

Pass interval as second argument when running:

./metrics_collector http://localhost:8080/metrics 10  # Collect every 10 seconds

API Reference

Go Server Endpoints

POST /metrics

Submit new metrics:

{
  "hostname": "server-01",
  "timestamp": 1234567890000000,
  "cpu_usage": 45.2,
  "memory_usage": 62.8,
  "disk_usage": 34.5,
  "disk_io_read": 1024000
}

GET /query

Query historical metrics:

• ?hostname=server-01 - Filter by hostname
• ?hours=6 - Get last 6 hours of data

Response:

[
  {
    "hostname": "server-01",
    "timestamp": 1234567890000000,
    "cpu_usage": 45.2,
    "memory_usage": 62.8,
    "disk_usage": 34.5,
    "disk_io_read": 1024000
  }
]

GET /latest

Get the most recent metrics for all hosts:

{
  "server-01": {
    "hostname": "server-01",
    "timestamp": 1234567890000000,
    "cpu_usage": 45.2,
    "memory_usage": 62.8,
    "disk_usage": 34.5,
    "disk_io_read": 1024000
  }
}