Darto Logo


darto_rate_limit

Distributed RateLimitStore backends for the Darto framework's core rateLimit() middleware — share the same counter across every app instance.

The core ships with an in-process MemoryRateLimitStore (zero-dep, perfect for a single instance). This package adds RedisRateLimitStore so the same limit applies when you're running multiple replicas behind a load balancer.

Install

dependencies:
  darto: ^1.2.0
  darto_rate_limit: ^1.0.0

Usage

import 'package:darto/darto.dart';
import 'package:darto/rate_limit.dart';
import 'package:darto_rate_limit/darto_rate_limit.dart';

void main() async {
  final store = await RedisRateLimitStore.connect(
    host: 'localhost',
    port: 6379,
    prefix: 'rl:',
  );

  final app = Darto()
    ..use(rateLimit(
      max: 100,
      window: Duration(minutes: 1),
      store: store, // ← all instances share this counter
    ));

  app.listen(3000);
  await store.close();
}

How it works

Each hit(key) runs a tiny Lua script in a single round-trip, so the counter is correct under concurrent hits from many app instances:

local n = redis.call('INCR', KEYS[1])
if n == 1 then redis.call('PEXPIRE', KEYS[1], ARGV[1]) end
return { n, redis.call('PTTL', KEYS[1]) }
  • INCR is atomic — counts are never lost.
  • PEXPIRE only fires when the window starts, so every instance agrees on the same resetAt.
  • One round-trip per request — no race, no extra latency vs the memory store.

Tests

dart test --tags redis

The included suite boots a disposable redis:latest container on a random host port (docker must be available).



Support 💖

If you find Darto Rate-Limit useful, please consider supporting its development 🌟Buy Me a Coffee.🌟

Libraries

darto_rate_limit
Distributed RateLimitStore backends for Darto.