<b>The forum that ate a 90Gbps flood and stayed up</b>

<b>The forum that ate a 90Gbps flood and stayed up</b>

A gaming forum on a single 1Gbps origin in Amsterdam got hit with a 90Gbps UDP and HTTP flood after a moderation drama. Their pipe saturated in seconds; the site was dark for 40 minutes before they could even reach the host.

We moved DNS behind an anycast CDN and locked the origin firewall to only accept traffic from the CDN's IP ranges. The next attack — 120Gbps — landed on the CDN's distributed edge, not the origin. Anycast spread it across dozens of points of presence, each absorbing a slice well under its capacity.

Real users in Berlin saw zero downtime. Origin bandwidth during the attack stayed flat at normal levels because the flood never reached it. Mitigation was automatic; nobody got paged.

The number that mattered: 120Gbps absorbed, 0 seconds of downtime.