So June 19th, 14:47 UTC. My monitoring goes red. Hetzner dashboard shows 600 Gbps inbound. On a 256 MB NAT VPS. $2.50/month.
I expected death. Instead: 3% packet loss for 8 minutes, then clean. My site stayed up.
I thought I was a genius. Emailed everyone. Then I read the incident report.
Contabo, my upstream, has an anycast scrubbing center in Amsterdam. They auto-activated for my IP prefix. I wasn't protected—I was collateral. The attack hit their range, they scrubbed, I got clean pipe.
The real trick? I paid Hetzner €3/month for "nullroute insurance" in their Falkenstein location. If Amsterdam failed, they'd nullroute me and refund 10x. They never triggered it. I just got lucky topology.
Specs:
- 256 MB RAM / 1 vCPU (KVM)
- 10 GB SSD (SATA, not NVMe)
- 1 IPv4 NAT + /64 IPv6
- 1 TB bandwidth (irrelevant at 600 Gbps)
Total outlay: $5.50/month. Uptime: 100%. Ego: destroyed.