docker-time-analysis

Statistical analysis tool for Docker container startup times using regression modeling and interactive visualizations.

Features

• CSV file upload and parsing
• Statistical analysis (correlation tables, descriptive statistics)
• Three regression model types: Linear, Exponential, Piecewise
• Statistical significance testing (R², Residual Variation, F-Stat Test, Pearson Coefficient)
• Interactive charts and visualizations with ECharts

Collected Data

The data provided in /data was collected inside a virtual machine (VMWare) running Ubuntu Server 24.04 LTS with the following configurations:

• Intel Core i5-10300H 4 Cores
• 8 GB RAM
• 80 GB SSD

System degradation at around 1000 Alpine Linux containers can be explained by resource exhaustion to create and/or sustain Docker containers.

Users are free to collect their own data by modifying/running /data/collect_data.sh

Screenshots

Overview of the application

Visualization of regression models

Regression models data

Regression models validation

Tech Stack

• Backend: Go, Gonum (statistical computing)
• Frontend: Svelte, TypeScript, ECharts

Project Structure

backend/
├── cmd/              - Entry points
├── internal/
│   ├── parse/        - CSV parsing and data validation
│   ├── compute/      - Statistical analysis and regression models
│   └── server/       - HTTP API server
frontend/
├── src/              - Svelte components and TypeScript logic
├── dist/             - Built output
└── vite.config.ts    - Build configuration
data/                 - Sample CSV data files

Getting Started

Prerequisites

• Go 1.26.1 or later
• Node.js 18+ with npm

Run as Docker Compose

git clone https://github.com/ItakawaM/docker-time-analysis
docker-compose up -d

Run Manually

Backend:

cd backend
go build -o server.exe ./cmd
./server.exe

or

cd backend
make build
./server.exe

Frontend:

cd frontend
npm install
npm run dev    # Development server
npm run build  # Production build

The server runs on http://localhost:8080 by default.

Requirements

• Go: 1.26.1+
• Node.js: 18+

License

MIT License — see LICENSE