Skip to main content

High Level Design

Rate-Limiter as a Middleware​

Redis is used as an in-memory key-value database for fast access to counters, making rate-limiting checks very low latency.

Rate-Limiter-1

Deep-dive: Rules​

Question: How do we configure and manage rate-limiting rules?

The rules (limits) are stored in a persistent database (Rules DB). Workers periodically check for any changes in these rules. If an update is detected, the new rules are processed and pushed to an in-memory cache (Rules Cache).

The rate limiter then reads from this Rules Cache to enforce limits in real time.

Rate-Limiter-rules-1

For excalidraw file, click here to download