Small toolkit for diagnosing unstable network behavior at work or home.

• speedtest.py: HTTP upload test with live latency sampling and JSON + HTML reports.
• report_to_html.py: Convert a JSON report into a self-contained interactive HTML plot.
• server/: Minimal Python upload sink (/up, /health) plus Dockerfile for self-hosting.

python3 speedtest.py --upload-endpoint https://speed.cloudflare.com/__up --report
python3 speedtest.py --upload-endpoint http://127.0.0.1:8080/up --upload-size 40 --upload-limit 1 --report

--report now exports both <timestamp>.json and <timestamp>.html.

cd server
docker compose up --build

Then point speedtest.py upload endpoint to your own server URL.

python3 report_to_html.py 2026-03-03_11-36-52.json

This writes 2026-03-03_11-36-52.html next to the JSON file. You can open it by double-clicking the HTML file in your file manager.

If you want a lower-level throughput baseline, run iperf3 between two hosts:

# on the remote/server host
iperf3 -s

# from your test machine (upload-like test: client -> server)
iperf3 -c <server-host-or-ip> -t 30 -i 1 -P 4

# from your test machine (download-like test: server -> client)
iperf3 -c <server-host-or-ip> -t 30 -i 1 -P 4 -R

-i 1 prints per-second stats, -P 4 uses parallel streams, and -R reverses direction.

• Public test endpoints may timeout or rate-limit long uploads.