How Bots Bypass Rate Limits
in Real-World APIs
If you're relying on `rate_limit_per_ip` to keep your API safe, you're leaving the door wide open. Here's how attackers circumvent your limits and what you can do about it.
1. IP Rotation (Low and Slow)
The most common technique is using a massive pool of residential proxies. By rotating through thousands of IPs, a bot can hit your endpoint 100,000 times while each individual IP only sends 2-3 requests—staying well below any standard rate limit threshold.
2. Header Spoofing
Bots often spoof `X-Forwarded-For` or `CF-Connecting-IP` headers to trick poorly configured load balancers into rate limiting the wrong IP, or worse, the load balancer's own internal IP.
3. Distributed State Attacks
Attackers now coordinate across different providers. Some traffic comes from AWS (the "recon" bots), while the "attack" bots use residential networks. They spread the load so perfectly that your logs look like a spike in organic traffic.
Beyond the IP: Use Signal Multi-Tenancy
To stop these attacks, you need to look beyond the IP address. You need to analyze the **ASN Governance**, the **Connection Type**, and the **Behavioral Intent** of the request.
Sentinel tracks "Global Velocity" across our entire network. If an IP rotated out of an attack on one platform and hits your API 30 seconds later, we already know it's Untrusted.
Secure Your API