KCNA — Mock Exam Part 2
Mock Exam Part 2: Cloud Native Architecture, Observability & App Delivery
15 questions covering KCNA Domains 3, 4, and 5 (32% of the real exam). Time yourself — aim for 22 minutes (1.5 min/question).
Question 1
Which of the following BEST describes “cloud native”?
- A) Any application running in the cloud
- B) Applications built using microservices, containers, service meshes, immutable infrastructure, and declarative APIs
- C) Applications written in Go or Rust
- D) Applications that only run on Kubernetes
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Simply running in the cloud doesn’t make an app cloud native — lift-and-shift is not cloud native |
| B | The CNCF definition: cloud native technologies empower organizations to build and run scalable applications using containers, microservices, service meshes, immutable infrastructure, and declarative APIs |
| C | Language choice doesn’t determine cloud-native status |
| D | Cloud native is a design philosophy, not a platform requirement |
Question 2
What does “immutable infrastructure” mean?
- A) Infrastructure that cannot be created
- B) Infrastructure that is never modified after deployment — replaced instead
- C) Infrastructure that uses only read-only file systems
- D) Infrastructure that doesn’t change its IP address
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Immutable infrastructure is created — it just isn’t modified after creation |
| B | Immutable infrastructure means servers/containers are never patched or modified in place. To update, you build a new image, deploy it, and destroy the old one. This eliminates configuration drift. |
| C | Read-only file systems are a security practice, not the full concept |
| D | IP address stability is unrelated to immutability |
Question 3
Which of the following is a CNCF Graduated project?
- A) Docker Desktop
- B) Amazon ECS
- C) Prometheus
- D) Azure Monitor
Show Answer
Correct Answer: C
| Option | Explanation |
|---|---|
| A | Docker Desktop is a Docker Inc. product, not a CNCF project |
| B | Amazon ECS is an AWS proprietary service |
| C | Prometheus is a CNCF Graduated project (graduated 2018). It’s the de-facto standard for cloud native monitoring. |
| D | Azure Monitor is a Microsoft proprietary service |
Question 4
In a microservices architecture, how do services typically communicate?
- A) Shared memory only
- B) Direct function calls
- C) Lightweight protocols like HTTP/REST or gRPC, and asynchronous messaging
- D) By reading each other’s databases
Show Answer
Correct Answer: C
| Option | Explanation |
|---|---|
| A | Shared memory is for inter-process communication within one host |
| B | Direct function calls imply monolithic or tightly coupled architecture |
| C | Microservices communicate via lightweight APIs (HTTP/REST, gRPC) for synchronous calls, and message brokers (Kafka, NATS, RabbitMQ) for asynchronous communication |
| D | Each microservice should own its data — accessing another service’s database is an anti-pattern |
Question 5
What are the three pillars of observability?
- A) Logs, Metrics, Traces
- B) Monitoring, Alerting, Dashboards
- C) CPU, Memory, Network
- D) Availability, Latency, Throughput
Show Answer
Correct Answer: A
| Option | Explanation |
|---|---|
| A | Logs (discrete events), Metrics (numeric measurements over time), and Traces (request paths through distributed systems) are the three pillars of observability |
| B | Monitoring, alerting, and dashboards are practices built on top of the three pillars |
| C | These are resource types, not observability pillars |
| D | These are performance metrics, not the observability framework |
Question 6
Prometheus uses a ______ model to collect metrics.
- A) Push
- B) Pull (scraping)
- C) Streaming
- D) Polling via agents
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Push-based systems (like Graphite, InfluxDB) require targets to send metrics to the server. Prometheus uses Pushgateway only for short-lived jobs. |
| B | Prometheus uses a pull (scraping) model — it periodically scrapes HTTP endpoints (e.g., /metrics) on targets. This lets Prometheus control the collection rate and detect target failures. |
| C | Streaming is used by systems like Kafka, not Prometheus’s primary model |
| D | Prometheus doesn’t require agents — targets expose metrics endpoints directly |
Question 7
What does an SLO (Service Level Objective) define?
- A) The contractual penalties for downtime
- B) A target value or range for a service level measured by an SLI
- C) The maximum cost of running a service
- D) The team responsible for a service
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Contractual penalties are part of SLAs (Service Level Agreements) |
| B | An SLO is a target — e.g., “99.9% of requests complete in under 300ms.” It’s measured by SLIs (Service Level Indicators) and is stricter than the SLA. |
| C | Cost management is FinOps, not SLOs |
| D | Team responsibility is organizational, not an SLO |
Question 8
In GitOps, what is the “single source of truth”?
- A) The container registry
- B) The Kubernetes cluster state
- C) A Git repository containing declarative configuration
- D) The CI/CD pipeline
Show Answer
Correct Answer: C
| Option | Explanation |
|---|---|
| A | Container registries store images, not desired cluster state |
| B | The cluster state is the actual state — it’s reconciled against the desired state in Git |
| C | In GitOps, the Git repository is the single source of truth. All desired infrastructure and application state is declared in Git. An operator (ArgoCD/Flux) reconciles the cluster to match Git. |
| D | The CI pipeline builds and tests — the repository is the source of truth |
Question 9
What does Helm do in the Kubernetes ecosystem?
- A) Manages container images
- B) Provides a package manager for Kubernetes applications
- C) Replaces kubectl
- D) Monitors cluster health
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Container images are managed by registries (Docker Hub, ECR, GCR) |
| B | Helm is the package manager for Kubernetes. It bundles K8s resources into charts, supports templating with values, manages versioned releases, and simplifies deploying complex applications. |
| C | Helm works alongside kubectl, not as a replacement |
| D | Monitoring is done by Prometheus, Datadog, etc. |
Question 10
Which deployment strategy runs two identical production environments?
- A) Rolling update
- B) Canary
- C) Blue-green
- D) Recreate
Show Answer
Correct Answer: C
| Option | Explanation |
|---|---|
| A | Rolling updates gradually replace pods — single environment |
| B | Canary routes a small percentage of traffic to the new version alongside the old |
| C | Blue-green maintains two identical environments — Blue (current) and Green (new). Traffic is switched from Blue to Green once validated. Instant rollback by switching back. |
| D | Recreate kills all old pods before creating new ones — single environment |
Question 11
What is the Twelve-Factor App methodology primarily about?
- A) Building mobile applications
- B) Best practices for building modern, portable, scalable SaaS applications
- C) A security framework for containers
- D) A Kubernetes resource specification
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Twelve-Factor is platform-agnostic but especially relevant to web/SaaS apps |
| B | The Twelve-Factor App methodology defines 12 principles for building modern SaaS applications: codebase, dependencies, config, backing services, build-release-run, processes, port binding, concurrency, disposability, dev/prod parity, logs, admin processes |
| C | It’s a development methodology, not a security framework |
| D | It predates Kubernetes — it’s an application design methodology |
Question 12
In distributed tracing, what is a “span”?
- A) The total time of a request
- B) A single unit of work within a trace, representing one operation
- C) The network latency between two services
- D) A log entry with a timestamp
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | The total time of a request is represented by the root span / trace duration |
| B | A span represents a single operation within a trace — it has a name, start time, duration, and metadata. Spans have parent-child relationships that form a trace tree. |
| C | Network latency is one aspect that might be visible in spans, but a span is the entire operation |
| D | Log entries are separate from spans — though they can be correlated |
Question 13
What is ArgoCD primarily used for?
- A) Building container images
- B) Running automated tests
- C) Continuous delivery using GitOps principles
- D) Managing cloud infrastructure billing
Show Answer
Correct Answer: C
| Option | Explanation |
|---|---|
| A | Image building is done by tools like Docker, Buildah, or Kaniko |
| B | Test automation is part of CI pipelines (Jenkins, GitHub Actions) |
| C | ArgoCD is a declarative GitOps continuous delivery tool for Kubernetes. It syncs cluster state with Git repositories, providing automated deployments, drift detection, and rollback. |
| D | Billing management is a FinOps function |
Question 14
What does “serverless” mean in the context of cloud computing?
- A) There are no servers involved
- B) The cloud provider manages server infrastructure; developers focus only on code
- C) Applications run on the client side only
- D) Servers are replaced by databases
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Servers still exist — they’re just abstracted away from the developer |
| B | Serverless means the cloud provider provisions, scales, and manages the infrastructure. Developers deploy functions or containers without managing servers. Examples: AWS Lambda, Google Cloud Functions, Knative on K8s. |
| C | Serverless apps run on cloud infrastructure, not client-side |
| D | Databases don’t replace servers — serverless databases exist alongside serverless compute |
Question 15
What is the primary difference between CI and CD?
- A) CI is for testing, CD is for monitoring
- B) CI automates building and testing code; CD automates releasing and deploying code
- C) CI uses Git, CD uses Docker
- D) CI is for frontend, CD is for backend
Show Answer
Correct Answer: B
| Option | Explanation |
|---|---|
| A | Monitoring is separate from CI/CD |
| B | CI (Continuous Integration) automates building, testing, and merging code changes. CD (Continuous Delivery/Deployment) automates releasing validated code to staging/production. Together they form the CI/CD pipeline. |
| C | Both CI and CD can use Git and Docker — they’re not tool-specific |
| D | CI/CD applies to all application types regardless of layer |
Score Yourself
| Score | Rating | Action |
|---|---|---|
| 13-15 | Excellent | You’re well-prepared for Cloud Native domains |
| 10-12 | Good | Review topics you missed |
| 7-9 | Fair | Go back to Chapters 5-7 and reinforce weak areas |
| Below 7 | Needs Work | Re-study Chapters 5-7 before the exam |
Combined Score (Part 1 + Part 2)
| Score (out of 30) | Rating | Readiness |
|---|---|---|
| 25-30 | Excellent | You’re exam-ready! |
| 20-24 | Good | Minor review needed |
| 15-19 | Fair | Focused study required |
| Below 15 | Needs Work | Comprehensive review recommended |
What’s Next?
Complete the final checklist to make sure you’re fully prepared for exam day.
