All DDoS Definitions
DDoS Testing Definition

DDoS War Game

When the alerts on the dashboard are real, the provider's phone line is answered by a real engineer, and the on-call team has to decide with traffic still arriving, nobody is reading from a script any more. That is a DDoS war game: a live, authorised exercise in which mixed attack traffic is sent at in-scope systems while the team that would respond to a real attack responds to this one. It is the form of DDoS test that measures the people and the controls in the same hour, rather than one without the other.

DDoS war game: real traffic, real responders, one clock Authorised traffic mixed vectors, several tiers ceilings and window agreed In-scope systems edge, provider, origin real alerts, real dashboards signals Responders on call, real runbook real access, real channels Exercise control holds scope and the stop decision, does not help Observers time each step on one clock, answer nothing What it produces Two-track timeline traffic events beside team decisions Control results per vector which held, which needed a person to act Gaps between them unread alerts, missing access, runbook steps left behind BlackNeuron™
DDoS war game: an exercise control team holding the stop authority sits between authorised mixed attack traffic and the on-call responders, observers time each step on one clock, and the outputs are a timed response log, control results per vector and the gaps between them

How a DDoS war game is run

Three groups take part, and their roles do not overlap.

Exercise control. A small group that holds the authorisation, the agreed scope and the stop decision. It watches the targets and the business service, and it can end the exercise at once. It does not help the responders.

The responders. The on-call engineers, the network and platform owners, the incident lead, and whoever speaks to customers. They work from the real DDoS runbook, with their real access, through their normal channels.

Observers. People who time each step on one clock and write down what the responders did, saw and asked for. They do not answer questions.

The traffic is a set of vectors aimed at different tiers, often overlapping, the way a multi-vector attack arrives. The vectors, targets, rate ceilings and time window are written down and authorised before the day. The responders may or may not know the date; everyone who has to authorise the traffic always does.

Why it matters under attack

The decisions are made on real signals. In a discussion exercise, the facilitator hands out the facts. In a war game, the team has to find them: which dashboard moved first, whether the provider's alert arrived, whether the logs are readable while the link is full. Many delays come from that search, not from the decision itself.

A second vector tests attention. Teams under pressure tend to fix the first thing they see and stop looking. When a second vector starts while the first is still being handled, the war game shows whether anyone notices, and how late.

Every hand-off is exercised for real. The call to the provider, the request to change a rule, the approval to block a region. Each depends on a person being reachable and holding the right permission today, not when the runbook was last edited.

It finds the gap between the tiers. A control can pass its own test and the response can still fail, because the alert went to a mailbox nobody reads. A war game is where that kind of gap shows up, since it is neither purely technical nor purely procedural.

DDoS war game: the team and the traffic on one clock normal service before the window opens service health exercise window, timed by observers Exercise opens Vectors land: was it declared? Second vector, rule changed Recovered: log both tracks Exercise control holds the stop decision throughout; every response step is timed against the traffic BlackNeuron™
Animation of a DDoS war game on one clock: service health holds as the exercise opens, falls when the vectors land and the team must declare, starts to recover once a second vector is spotted and a rule is changed, and returns to normal with both tracks logged

What a DDoS war game measures

The main output is a timeline with two tracks. One is the traffic: when each vector started, when the edge or the provider began to mitigate it, when service recovered. The other is the team: when the incident was declared, when each decision was taken, when each action took effect.

Read side by side, they show where the time to mitigation went. A long gap between detection and declaration is a people problem. A long gap between the decision and the change taking effect is a tooling or permission problem. A long gap with no human step in it is a control problem.

Alongside the timeline, the observers' notes record every point where the team left the runbook, asked for something it did not have, or could not reach someone. Those go into the DDoS test report with the technical results, each with an owner.

A war game is expensive in people's time, so it is worth running once the controls already pass their own tests and the decisions have been rehearsed on paper first. That rehearsal, with no traffic at all, is what a DDoS tabletop exercise is for; the war game is where its assumptions get timed.