Chapter 3

Measure → Change → Measure

The only legitimate way to optimise.

~14 min total2 lessonsAssignment

By the end of this chapter you'll be able to

Use benchstat to compare runs
Make one change and prove the win or roll back

Lessons

Module: Disciplined optimisation

benchstat — compare two runs

Statistical significance, not vibes.

6mmedium

Ship one measured optimisation

Pick, change, re-run, decide.

8mmedium

Chapter assignment

One real, measured win

Make ONE change to your service's hot path. Re-run benchmarks. Use benchstat to verify the change is real (p<0.05). If not, roll back. Paste before/after in notes.md.

See success criteria

Lesson 1 takes ~6 min.

Start chapter

Spot a typo? Have an idea?

Help us improve this chapter. One click opens a GitHub issue with this chapter's URL pre-filled — suggest a clearer wording, report a bug, or request a new lesson. The course keeps improving thanks to learners like you.

Suggest an improvement on GitHub

ch.2: Profiling with pprof