On Sunday 26 November 2023 at 10:57 UTC, a memcached amplification targeted a online casino customer in Southeast Asia. The attack peaked at 265.6 Gbps and lasted 122 minutes. Traffic originated from 2279 autonomous systems in 69 countries, predominantly a rented booter service.
| Vector | memcached amplification |
| Peak | 265.6 Gbps |
| Duration | 122 min |
| Time to mitigation | 0.481 s |
| Attack traffic reaching origin | 0.022% |
| Legitimate traffic challenged | 0.56% |
Timeline
The attack was preceded by a hacktivist channel announcing the target. Edge packet filters identified the flood by source port and payload signature and dropped it at line rate across 27 points of presence.
What the customer saw
A brief increase in p99 latency of 182 ms during the first minute, then normal service.
Recommendations
- Review allow-listed partner ranges quarterly.
- Add a dedicated rate limit for the targeted route.
- Keep origin IPs out of public DNS history.
Clear and practical, thanks.
Thanks! Yes — the risk score and its components are included in every log record.
Great write-up. We saw almost the same pattern on our login endpoint last month.
Verified good bots and allow-listed partners bypass challenges entirely.
Interesting that most attacks are under ten minutes. Our experience is similar.
Is the risk score exposed in the logs so we can build our own dashboards on it?
Good question. We will cover that in a follow-up post.
We had the exact false-positive issue with CGNAT carriers. The weighting change makes sense.
Would love a follow-up on how you handle HTTP/3 fingerprinting.