Latency Test: TCP Ping Between Endpoints

This project measures TCP round-trip latency from any Linux host to a specified endpoint using TCP ping.

Features

• Measures TCP round-trip latency to a specified endpoint and port.
• Logs destination IP address for each request.
• Results saved in CSV format for analysis.
• Each run produces a uniquely named results file with the UNIX timestamp.
• The last row of the CSV contains the average response time and the start time UNIX timestamp.

Requirements

• Python 3.8+
• traceroute
• Docker (optional)
• No external Python dependencies required for the TCP ping script.

Usage

1. Direct Python Execution

Install dependencies (Alpine example):

apk add python3 traceroute
mkdir -p results

Download tcpping script and run:

wget https://raw.githubusercontent.com/deajan/tcpping/master/tcpping
chmod +x tcpping
export TCP_ENDPOINT="your-endpoint"
export TCP_PORT=443
export NUM_REQUESTS=10
python3 script.py

Results will be saved in the results/ directory.

2. Using Docker

Build the Docker image:

docker build -f Dockerfile -t latency-tcp .

Or use the pre-built image from GitHub:

The container image is available at ghcr.io/peanutsguy/tcpping_latency.

Run the container:

docker run --rm \
  -e TCP_ENDPOINT="your-endpoint" \
  -e TCP_PORT=443 \
  -e NUM_REQUESTS=10 \
  -v $(pwd)/results:/app/results \
  ghcr.io/peanutsguy/tcpping_latency

Results will be available in your local results/ directory.

Output

CSV file columns:

• request
• latency_ms
• destination_ip
• tcp_port
• msg (error or status or hostname)
• start_time (only in the summary row)

The last row contains:

• request: "Average"
• latency_ms: average response time (ms)
• start_time: UNIX timestamp when the test started
• msg: hostname

Security

• Do not commit sensitive credentials.
• Use environment variables for endpoint configuration.

References

• tcpping GitHub