CNPA Delivery, APIs, and Observability Review
CNPA Track | Multiple-choice exam prep | Delivery, APIs, observability, and measurement
Why This Module Matters
Section titled “Why This Module Matters”This module covers the parts of CNPA that prove a platform is not just a philosophy. A good platform has delivery flows, provisioning APIs, observability, policy, and metrics that tell you whether the platform is actually helping.
If the previous module was about the platform as a product, this one is about how the product behaves in practice.
What You’ll Learn
Section titled “What You’ll Learn”After this module, you will be able to:
- distinguish delivery, release, and promotion concepts
- explain why reconciliation-based provisioning matters
- identify observability building blocks and conformance controls
- choose the right high-level metric for platform success
- avoid exam traps around tooling, API style, and feedback loops
Part 1: Delivery And Release
Section titled “Part 1: Delivery And Release”CNPA questions about delivery usually test whether you know the difference between:
- continuous integration
- continuous delivery
- continuous deployment
- progressive delivery
- release promotion
The exam does not expect implementation details. It expects you to understand that a platform should make delivery repeatable, observable, and safe.
Common distinctions
Section titled “Common distinctions”| Term | Meaning |
|---|---|
| Continuous integration | Changes are merged and verified frequently |
| Continuous delivery | The system is always in a deployable state |
| Continuous deployment | Deployments happen automatically after checks pass |
| Progressive delivery | Release strategies such as canary or blue-green |
| Promotion | Moving the same artifact across environments with controls |
The right answer is often the one that says “reduce manual handoffs and make outcomes reproducible.”
Part 2: Platform APIs And Provisioning
Section titled “Part 2: Platform APIs And Provisioning”The exam often uses API language to test whether you understand self-service infrastructure.
Good platform APIs:
- are declarative
- return a stable desired state
- hide implementation details behind a clear contract
- can be reconciled repeatedly
This is where CRDs, operators, and Crossplane-style abstractions matter. The point is not “Kubernetes everywhere.” The point is exposing a safe, reusable interface to platform consumers.
Recommended study anchors:
What The Exam May Ask
Section titled “What The Exam May Ask”- What makes a provisioning API “platform-like” rather than “ticket-like”?
- Why is reconciliation useful for infrastructure requests?
- How do CRDs fit into self-service infrastructure?
- What is the relationship between claims, composite resources, and the underlying provider resources?
Part 3: Observability And Conformance
Section titled “Part 3: Observability And Conformance”Observability is not the same as logging. CNPA wants you to understand the full picture:
- metrics show trends
- logs show events and context
- traces show request flow across services
Conformance is about proving the platform behaves consistently and safely. That can include:
- policy enforcement
- standards for app delivery
- security guardrails
- reporting on whether workloads meet the expected platform baseline
Recommended study anchors:
- What is Observability?
- The Three Pillars
- From Data to Insight
- Security Mindset
- OPA & Gatekeeper
- Kyverno
Observability Traps
Section titled “Observability Traps”| Trap | Better answer |
|---|---|
| Monitoring and observability are synonyms | Monitoring is predefined; observability supports new questions |
| High-cardinality data is always good | High-cardinality data is useful in the right telemetry layer, not as a blanket metric label strategy |
| Policies are only for security teams | Platform policies are part of the user experience and platform baseline |
Part 4: Measuring Your Platform
Section titled “Part 4: Measuring Your Platform”This domain is small, but it is not soft. The exam wants to know whether you understand that a platform should be measured like a product.
Useful measurement categories:
- adoption
- developer satisfaction
- time to first successful deploy
- time to provision a standard service
- error rate or incident reduction
- policy compliance or conformance
SRE concepts help here:
The point is to prove the platform improves outcomes, not just that it exists.
Common Mistakes
Section titled “Common Mistakes”| Mistake | Why it hurts | Better answer |
|---|---|---|
| Treating delivery as only “CI/CD” | CNPA also cares about promotion and safe release | Include release strategy and environment flow |
| Forgetting observability is a system property | The exam may separate telemetry from insight | Explain the system, not just the tools |
| Overfocusing on one metric | A platform can look busy and still be ineffective | Use adoption plus reliability plus developer outcomes |
| Treating APIs as implementation details | The platform itself is often exposed through APIs | Think self-service contract first |
Mini Recall Check
Section titled “Mini Recall Check”- Can you explain why continuous deployment is not the same as progressive delivery?
- Can you describe a self-service API without naming a product?
- Can you separate observability from monitoring in one sentence?
- Can you name at least two metrics that would tell you whether a platform is adopted?
Next Module
Section titled “Next Module”Continue with CNPA Practice Questions Set 1.