11.2. Zero Trust Architecture
Zero trust replaces the perimeter model — trust the internal network, guard the edge — with a single principle: never trust, always verify, regardless of where a request originates. Every request is authenticated, authorized, and encrypted as if it came from the open internet, and access decisions are made continuously rather than once at login. This chapter covers the engineering of that principle: micro-segmentation, identity-aware proxies, Google’s BeyondCorp model, continuous verification, and policy-as-code that makes access rules auditable and version-controlled.
Topics Covered
Section titled “Topics Covered”- 11.2.1. Engineering the Never Trust, Always Verify Principle: Engineers the never-trust-always-verify principle: authenticate and authorize every request regardless of origin.
- 11.2.2. Micro-Segmentation: The Least-Privilege Network Model: Covers micro-segmentation, enforcing least-privilege network access between individual workloads.
- 11.2.3. Identity-Aware Proxy (IAP): Explains the identity-aware proxy that authorizes each request based on user and device identity.
- 11.2.4. The BeyondCorp Model: Google’s Approach: Covers Google’s BeyondCorp model that moved trust from the network perimeter to authenticated identity.
- 11.2.5. Continuous Verification: Dynamic Security Decisions: Explains continuous verification, re-evaluating trust dynamically rather than once at login.
- 11.2.6. Policy as Code: OPA, Gatekeeper, and Kyverno: Covers policy as code with OPA, Gatekeeper, and Kyverno for auditable, version-controlled access rules.