Lesson 24 — Concurrency limiter

Only N at a time

Firing 100 requests at once can overload a server or hit a rate limit. A concurrency limiter runs async tasks with at most N in progress: start N workers, and each worker keeps taking the next task until none are left. Results are stored by index so they come back in the original order.

// 2 workers, 3 tasks: worker A takes task 0, worker B takes task 1,
// whoever finishes first takes task 2.

Why it matters: this protects APIs and memory in scrapers, uploads and batch jobs — a real interview and production pattern.