A high-performance distributed rate limiter based on the Token Bucket algorithm, designed for Node.js applications using Redis as the backing store.
-
Token Bucket rate limiting
-
Redis-backed distributed state
-
Atomic operations using Lua scripts
-
TypeScript support
-
Lightweight and framework-agnostic
-
Suitable for APIs, services, and microservices
npm install distributed-limiter
import { TokenBucketLimiter } from "distributed-limiter";
const limiter = new TokenBucketLimiter({
redis,
capacity,
refillRate
});
const result = await limiter.isAllowed({
key: "user:123",
tokenRequest: 1,
});
if (result.allowed) {
console.log("Request allowed");
} else {
console.log("Rate limit exceeded");
}Creates a new limiter instance.
const limiter = new TokenBucketLimiter({
redis,
limiterCapacity,
refillRate
});Checks whether a request can consume the requested number of tokens.
const result = await limiter.isAllowed({
key: string,
tokenRequest: number,
});key - Unique identifier (user, IP, API key, etc.)
capacity - Maximum bucket size
refillRate - Tokens added per second
tokenRequest - Tokens required for the request
Note: capacity and refillRate can be configured once during limiter initialization and reused automatically for all requests. In applications, only key and tokenRequest need to be supplied per request.
{
allowed: boolean;
}
Client Request
│
▼
BucketLimiter
│
▼
Redis Lua Script
│
▼
Token Bucket State
The limiter stores bucket state in Redis and performs all token calculations atomically through Lua scripts to ensure correctness across multiple application instances.
MIT