DDoS Attack

SECURITY

Quick Definition

A Distributed Denial of Service (DDoS) attack uses many compromised machines (a botnet) to flood a target with traffic. The goal is not to steal data; it is to make the service unavailable. DDoS attacks measure in packets per second (Pps) for network-layer floods or requests per second (Rps) for application-layer floods. The largest recorded attacks have peaked at 70+ million requests per second.

How it works

Network-layer DDoS (L3/L4) sends massive volumes of UDP or TCP packets to exhaust bandwidth or connection tables. Application-layer DDoS (L7) sends real-looking HTTP requests at a rate that overwhelms the application logic, often targeting expensive endpoints (search, login, checkout). Reflection attacks (DNS, NTP, memcached amplification) abuse misconfigured public servers to multiply traffic by 50-50,000x.

Defenses: anycast networks (Cloudflare, Akamai) absorb network-layer floods at the edge; rate limiting, JavaScript challenges, and bot fingerprinting handle application-layer attacks; circuit breakers and dependency isolation limit the blast radius when defenses do fail.

Why it matters

DDoS is now table-stakes risk for any internet-facing service. Renting a 10-Gbps DDoS for an hour costs single-digit dollars. Without a CDN or DDoS-protection layer, a small grudge can cost a service hours of downtime.

Where you'll see this on TerminalFeed

TerminalFeed runs behind Cloudflare, which absorbs DDoS at the edge. The service-status endpoint tracks availability of major DDoS-protection providers.