On Friday 21 May 2021 at 02:00 UTC, a HTTP/2 rapid reset targeted a gaming community customer in Latin America. The attack peaked at 948.9 million requests per second and lasted 187 minutes. Traffic originated from 1446 autonomous systems in 32 countries, predominantly a Mirai-derived IoT botnet.
| Vector | HTTP/2 rapid reset |
| Peak | 948.9 million requests per second |
| Duration | 187 min |
| Time to mitigation | 0.195 s |
| Attack traffic reaching origin | 0.061% |
| Legitimate traffic challenged | 0.08% |
Timeline
The attack was preceded by no stated motive. Request rates on the targeted routes exceeded their hourly baseline by a factor of 547 within 25 seconds. The risk score of participating clients crossed the challenge threshold automatically and proof-of-work difficulty rose with origin load.
What the customer saw
A brief increase in p99 latency of 90 ms during the first minute, then normal service.
Recommendations
- Keep origin IPs out of public DNS history.
- Add a dedicated rate limit for the targeted route.
- Review allow-listed partner ranges quarterly.
Nice to read a vendor blog that admits what went wrong.
Good question. We will cover that in a follow-up post.
The point about origin IPs leaking through certificate transparency logs is underrated.
We moved from a scrubbing provider to always-on last year; time to mitigation went from minutes to basically nothing.