Availability is a metric, the single number a DDoS test exists to defend and the outcome an attacker is trying to degrade. It measures the share of time a service responds correctly to legitimate users, usually expressed as a percentage (the familiar "nines": 99.9%, 99.99%) or as an error-and-latency budget over a window. A DDoS attack does not need to crash anything to win; pushing enough requests to inflate error rates or latency past a usable threshold is an availability failure even if the servers stay up.
Why it matters in DDoS testing
Raw throughput numbers can look healthy while real users are timing out, so availability has to be measured from the legitimate user's perspective, not the network's. A test tracks goodput and error rate for synthetic real-user traffic while the attack runs, capturing the exact load at which availability breaks rather than when the link fills.
For how this metric anchors a resilience baseline, see DDoS Resilience Testing.