TCP Congestion Control Comparison on a Long-Fat Network
Overview
What this challenge is about.
Set up two Linux test hosts in Sydney + Frankfurt cloud regions (or one host pair with tc-netem emulating 280ms RTT). Run iperf3 transfers using CUBIC and BBR at 4 loss rates (0.01 percent, 0.1 percent, 0.5 percent, 1 percent). Repeat each measurement 5 times with confidence intervals. Measure goodput, retransmission rate, and RTT inflation. Test with both single and 8 parallel streams. Deliver methodology, raw results, an 8-page report with per-condition recommendation and sensitivity analysis on loss-rate assumptions.
The Brief
What you'll do, and what you'll demonstrate.
Compare CUBIC and BBR TCP congestion controllers on a long-fat-network path across 4 loss rates and recommend a default for the storage fleet.
Earning criteria — what you'll demonstrate
- Compare TCP congestion controllers on long-fat networks rigorously
- Use tc-netem to emulate network conditions reproducibly
- Measure goodput, retransmission rate, and RTT inflation per controller
- Communicate infrastructure-default recommendations to a platform team
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Career mappings coming soon.