iPerf Data Analysis

This project provides tools to collect, analyze, and visualize data from iPerf3 network performance tests.

Project Structure

• config.R: Configuration settings
• data/: Raw and processed iPerf data
• R/: Functions for parsing, analyzing, and visualizing iPerf data
• scripts/: Main processing scripts
• output/: Generated results and visualizations
• data_collection/: Scripts for standardized data collection

Workflow

1. Collect Data

Use the provided data collection script:

cd data_collection
chmod +x run_iperf_test.sh
./run_iperf_test.sh

Options:

• -s SERVER: Server to test (can specify multiple times)
• -d SECONDS: Test duration in seconds (default: 120)
• -u USER: Username to record (default: current user)

Best Practices for Data Collection

Test Setup:

• Use geographically diverse servers for comprehensive testing
• Ensure test duration is at least 60 seconds (120s recommended)
• Minimize other network activity during tests
• Run tests at different times of day for better data variety

Recommended Command Options:

# Basic test
./run_iperf_test.sh -s speedtest.serverius.net -d 120

# Multiple servers
./run_iperf_test.sh -s speedtest.serverius.net -s speedtest.london.linode.com -d 120

See data_collection/BEST_PRACTICES.md for more detailed guidance.

2. Analyze Data

Copy the generated files to your input directory (set in config.R), then run:

source("scripts/process_all_data.R")

This will:

• Parse raw iPerf3 output files
• Clean and process the data
• Generate performance summaries
• Create visualizations

For customized analysis, you can also use individual scripts:

source("scripts/parse_raw_data.R")   # Convert raw data to processed format
source("scripts/generate_reports.R")  # Create summary reports
source("scripts/create_plots.R")      # Generate visualizations

3. View Results

Results are saved to:

• Summary reports: output/reports/
• Visualizations: output/figures/
• Processed data: data/processed/

Supported Formats

The analysis can process:

• Raw iPerf3 text output
• Pre-processed CSV files

Requirements

R packages:

• tidyverse
• fs
• patchwork
• scales
• zoo
• lubridate

Common Issues

1. Connectivity failures: The script attempts to ping servers first to avoid hanging.
2. Disk space: Long-term testing may accumulate large data files. Monitor available space.