Defence in Depth Vocabulary
Vocabulary Reference
| Term / Phrase | Meaning in context |
|---|---|
| Layers of security controls | Multiple independent controls addressing the same threat at different points — network, application, and data layers each having their own protections. |
| Compensating control | A control put in place to address a risk when the primary control is unavailable or insufficient — e.g. increased monitoring when patching is delayed. |
| "if one control fails" | Key design phrase expressing the defence-in-depth assumption that any single control may be bypassed — the architecture must not rely on a single control succeeding. |
| Security boundary | A defined point in a system where a trust transition occurs and security controls are enforced — e.g. the boundary between a public API and an internal service. |
| Blast radius | The extent of damage or access an attacker gains if a single control or component is compromised — defence in depth minimises blast radius at each layer. |
| Single point of failure (security) | A control whose failure would fully expose a system — defence in depth eliminates security single points of failure by providing independent fallback layers. |
1. The security architect explains the design: "If one control fails — say the WAF is bypassed — the application layer still validates all inputs, and even if that fails, the database uses parameterised queries." This design rationale illustrates:
2. A team cannot patch a critical vulnerability immediately due to a vendor delay. They implement enhanced monitoring and network isolation for the affected service in the meantime. The monitoring and isolation are:
3. The architecture diagram shows the line between the public-facing API gateway and the internal order processing service. At this line, TLS terminates, tokens are validated, and rate limits are enforced. What is this line called in security architecture terminology?
4. If an attacker compromises a single microservice, they gain access only to that service's data and cannot reach other services or databases. The design has limited the attacker's gain to a small part of the system. In defence-in-depth language, this limited gain is called:
5. The security review flags that the entire system relies on a single API key at the perimeter for authentication. "If this key is leaked, the entire system is exposed." The reviewer is identifying:
Exercise complete!
out of 5 questions