Rate Limiting
A control that bounds how quickly requests or operations are admitted over time to protect capacity, fairness or abuse-sensitive resources.
Admission Control
Rate limiting restricts the rate at which work is accepted. Token-bucket and leaky-bucket style mechanisms can permit controlled bursts while enforcing a longer-term rate.
The limit can be applied per identity, tenant, endpoint, source address or another policy key.
Capacity and Security
A rate limiter can reduce overload and abuse, but it does not increase downstream capacity. Limits should reflect the protected resource and should not turn one shared identity into a denial-of-service point for unrelated users.
Distributed implementations also need a clear consistency model.
Related Resilience Concepts
- Backpressure
- Circuit Breaker
- Retry
- Quota