Kubernetes and Cloud Native Associate Exam Prep

Practice progress

Practise the KCNA foundations

Unofficial, independently authored practice for the Linux Foundation / CNCF Kubernetes and Cloud Native Associate (KCNA). No affiliation or endorsement is implied.

Target: KCNA, English four-domain curriculum verified 13 September 2026. This is a local snapshot of an unnumbered outline, not a provider version or a tested Kubernetes minor.

Kubernetes Fundamentals 44%; Container Orchestration 28%; Cloud Native Application Delivery 16%; Cloud Native Architecture 12%. All 13 objectives are represented in 75 original questions.

The official exam is 60 multiple-choice questions in 90 minutes. These rounds are untimed original practice. Blueprint-weighted rounds follow the published domain weights, rounded to whole questions; they do not reproduce official questions or difficulty. No score predicts passing.

Official KCNA details · Official curriculum. Study links are for preparation; external resources are not allowed during the official multiple-choice exam.

Required curriculum verification date (optional):

Require the local snapshot used by your notes or saved link, or leave blank to use the available snapshot, 2026-09-13. This is not your exam date, a provider effective date or a live check for the latest curriculum.

Requested date: {{ JSON.stringify(params.curriculum_date) }}

URL date values: {{ JSON.stringify(conflict.values) }}

{{ params.curriculum_date.trim() ? 'The available snapshot meets this date requirement.' : 'No date requirement: use the disclosed available snapshot.' }}

Snapshot requirement shortcuts:

Both actions discard any saved attempt. Pin requires 2026-09-13; Use clears the date requirement. Start a new round after choosing.

Available snapshot

Practice mode:

Questions cover all four domains in the published proportions, rounded by largest remainder. Short rounds may omit a domain. The selected practice level limits the available counts.

Domain:

Practice difficulty:

Foundational concepts practise recognition and recall; applied reasoning uses scenarios. These authored levels are not calibrated to official exam difficulty.

Feedback:

Question count:

Replay seed:

Keep the seed to replay the same questions and option order.

Answers and the replay seed are saved in the page URL. Sharing that URL shares the attempt. No account, exam credentials or certification details are needed.

{{ progressPercent }}%

{{ answeredCount }} of {{ questionCount }} answered

Question {{ examCursor + 1 }} of {{ questionCount }}

Study: {{ studySource(currentQuestion.id).title }}

Answer recorded. Corrections appear after the round.

Review this practice round

Readiness review

Unofficial original KCNA practice, using the four-domain English curriculum snapshot verified 13 September 2026. This score describes only the authored questions in this round; it is not an official score or a passing prediction.

Revisit the linked documentation, practise the domains not sampled here, and build hands-on familiarity. This untimed concept bank does not reproduce official exam difficulty, proctoring or question selection, and does not replace the practical preparation needed for CKA, CKAD or CKS.

Missed answer review

{{ item.prompt }}

Your answer: {{ item.selected_answer_label }}

Correct answer: {{ item.correct_answer_label }}

{{ item.explanation }}

Study: {{ studySource(item.question_id).title }}

Domain performance

Unanswered or unsampled domains are not evidence of weakness.

{{ row.label }} · {{ row.weight_percent }}% blueprint weight: {{ row.question_count }} sampled; accuracy {{ row.accuracy_percent === null ? 'not sampled' : row.accuracy_percent.toFixed(1) + '%' }}; weighted contribution {{ row.contribution_points === null ? 'unavailable' : row.contribution_points.toFixed(1) + ' percentage points' }}.

Not tested this round: {{ untestedDomains.join("; ") }}.

Loading domain-performance chart…

Answer ledger

Question and domain	Answers	Explanation and study source	Copy
{{ row.question_id }} · {{ row.category_label }} · {{ itemDetails(row.question_id).objective }} · {{ difficultyLabel(row.question_id) }}{{ row.prompt }}	Your answer: {{ row.selected_answer_label }} Correct answer: {{ row.correct_answer_label }} {{ row.outcome_label }}	{{ row.explanation }} {{ studySource(row.question_id).title }}

A containerized application can start successfully and still be unreachable, unable to store data safely or unready to serve requests. Kubernetes and cloud native foundations explain these differences. They connect the application that runs inside a container with the scheduling, networking, storage and operational practices that support it.

Cloud native describes an approach to building and operating systems, not simply a choice of hosting provider. The Cloud Native Computing Foundation (CNCF) emphasizes repeatable automation, resilience, observability and systems whose parts can evolve without being tightly dependent on one another. These practices apply in public, private and hybrid clouds. Putting an existing application in a container is useful packaging, but does not automatically provide those operational properties.

Kubernetes coordinates containerized workloads across a cluster. Its control plane manages cluster state, while nodes provide the environment in which application containers run. Several closely related terms describe different responsibilities within that arrangement.

Container image: The application and its dependencies packaged for creating containers. An image can be stored and distributed without running the application.
Pod: Kubernetes' smallest deployable computing unit, containing one or more cooperating containers. Containers in a Pod share networking, but shared files require an appropriately mounted volume.
Deployment: A workload resource commonly used to maintain replicas of a stateless application and manage updates to their Pod template.
Service: A network abstraction that gives clients a stable way to reach backends even as individual Pods are replaced.

Declaring three replicas expresses a desired outcome. Controllers repeatedly compare that intention with what exists and work to restore the requested number of replicas. This process is called reconciliation. It does not guarantee that three healthy copies appear immediately: placement constraints, unavailable images or application startup failures can prevent the desired state from becoming reality.

These distinctions become useful when joining a team that deploys to Kubernetes, reviewing an application's operating requirements or investigating a failed rollout. More replicas do not repair a Service that selects the wrong Pods, and a running process does not prove that the application is ready for traffic. Understanding which mechanism is responsible helps identify the evidence to inspect before changing configuration.

Observability adds evidence about how the application behaves. Metrics provide numerical measurements such as request rate, logs record events and messages, and traces follow related work across a request's path. A rising error count can reveal a trend, while the reason for a particular failure may require a log message or a trace. The signals complement one another rather than serving as interchangeable answers.

The Kubernetes and Cloud Native Associate (KCNA) certification from The Linux Foundation and CNCF concerns foundational knowledge across Kubernetes and the wider cloud native ecosystem. Its scope includes application delivery, troubleshooting, observability and community collaboration as well as Kubernetes resource names. Conceptual preparation is useful for understanding these relationships, but answering questions does not demonstrate the ability to administer, debug or secure a production cluster.

The official KCNA exam has 60 multiple-choice questions with 90 minutes to complete them and is remotely proctored. External tools, resources and websites are not permitted during Linux Foundation multiple-choice exams. Documentation is therefore a preparation resource, not an exam-day reference allowance.

How to Use This Tool:

These are untimed, independently authored KCNA practice rounds, with no affiliation or endorsement from The Linux Foundation or CNCF. The initial setup uses Focused practice with 10 questions from All four domains, Mixed practice and feedback After each answer.

Set the optional Required curriculum verification date only when matching a snapshot used by your notes or saved link. Leave it blank to accept the available snapshot, or enter 2026-09-13. This identifies the local English curriculum snapshot, not your exam date, a provider effective date or a live check for updates.
A supplied date must be a real date in YYYY-MM-DD format, and any date other than 2026-09-13 blocks practice. For conflicting dates or an unavailable saved attempt, choose Pin available snapshot to require that date, or Use available snapshot to clear the requirement. Both actions discard the saved attempt; incompatible saved answers cannot be migrated.
Choose Practice mode. Focused practice lets you select one Domain or All four domains without fixed domain quotas. Blueprint-weighted practice allocates questions using the four published domain weights, rounded to whole questions.
Select Practice difficulty and an offered Question count. Foundational concepts emphasizes recognition and recall, Applied reasoning uses scenarios, and Mixed practice draws from both. Available counts depend on the selected pool; weighted rounds at either individual difficulty level offer at most 20 questions. These authored levels are not calibrated to official exam difficulty.
Choose when to see corrections under Feedback. After each answer reveals the correction and explanation immediately; After the round records the choice and postpones corrections until completion.
The optional Advanced section contains Replay seed. Keep the supplied seed or enter 1 to 128 characters. The same seed and setup reproduce the questions and option order while the question bank remains unchanged. An empty seed prevents the round from starting.
Select Start exam practice, then choose one of the four answers for each question. Selecting an answer records and locks it. Use Next question to continue and View results after the final answer. Reset and change setup discards the current attempt.
Review the explanations in Missed answer review and the complete Answer ledger. Use Retry missed to revisit only incorrect answers, Replay same round to repeat the same questions and option order, or New shuffled round to draw again from the available bank. Follow the study source for a missed concept before treating a successful retry as understanding.

Interpreting Results:

The headline fraction counts correct answers out of all questions in the completed round. Perfect round means that every answer in that round was correct. Start with the missed answers directs attention to errors. Neither label is a readiness level, an official pass decision or evidence of professional competence.

Domain performance needs to be read in the units shown. Focused rounds chart correct and missed answer counts. Weighted rounds chart contributions in blueprint percentage points, so a wider bar can reflect a larger domain weight rather than more questions. The separate Blueprint-weighted practice accuracy can differ from the percentage implied by the headline fraction.

Answered-domain coverage adds the weights of domains with at least one recorded answer. It does not measure correct answers, mastered objectives or the proportion of the question bank completed. Even 100% coverage can come from only a few questions. Domains listed under Not tested this round are unobserved, not demonstrated weaknesses.

Use the ledger's selected answer, correct answer, explanation and study source to identify the distinction you missed. Compare rounds with the same scope, difficulty, length and feedback timing, while allowing for differences in the questions sampled. Replays test familiar material, and new shuffled rounds can overlap earlier rounds. These percentages have no official score conversion or readiness threshold and predict no KCNA result.

Technical Details:

Kubernetes questions often turn on the difference between declaring an outcome, arranging the resources to achieve it and observing whether it actually happened. A valid resource declaration is not proof that an application is running or reachable. Each responsibility has a different mechanism.

Desired state, placement and execution

Kubernetes responsibilities and the limits of what each establishes
Responsibility	Mechanism	Important distinction
Accept cluster requests	The API server exposes the Kubernetes application programming interface (API).	Accepting a request does not run the application's business logic.
Store cluster data	etcd provides the backing key-value store for cluster data.	It is not the container image registry.
Reconcile workloads	Controllers act on differences between desired and observed state.	Restoring a replica count does not repair an application defect.
Place Pods	The scheduler selects suitable nodes for unassigned Pods.	Placement is separate from starting containers on the chosen node.
Run assigned Pods	The kubelet works with the container runtime on its node.	Running containers alone do not establish application readiness.

For objects that expose both fields, spec describes the intended configuration and status reports observed state. A Deployment's managing controllers normally replace a deleted Pod when the desired replica count remains unchanged. The replacement still needs a suitable node, an accessible image and successful startup. This explains why declarations, scheduling evidence and application evidence should not be treated as the same thing.

Health checks make another consequential distinction. A readiness probe determines whether a container is ready to receive normal Service traffic. A liveness probe can trigger a restart according to the applicable restart behavior. A startup probe delays readiness and liveness probing until initialization succeeds. Changing readiness alone does not prevent an independently failing liveness check from repeatedly restarting a slow-starting application.

Curriculum scope and question allocation

The practice target is the unnumbered, four-domain English KCNA outline represented by the local snapshot verified on 13 September 2026. That date is neither an official curriculum version nor a tested Kubernetes minor release. The bank declares concept coverage of the outline's 13 objectives through 75 original questions, with the following domain totals. These are original study questions, not questions taken from the official exam. Assignment to an objective is not exhaustive coverage of everything that objective can involve.

KCNA four-domain blueprint and the separate original practice-bank counts
Domain	Blueprint weight	Objectives	Original questions
Kubernetes Fundamentals	44%	Kubernetes Core Concepts; Administration; Scheduling; Containerization	33
Container Orchestration	28%	Networking; Security; Troubleshooting; Storage	21
Cloud Native Application Delivery	16%	Application Delivery; Debugging	12
Cloud Native Architecture	12%	Observability; Cloud Native Ecosystem and Principles; Cloud Native Community and Collaboration	9

Focused selection gives eligible questions an equal chance and samples without replacement. Choosing all four domains does not impose quotas on that focused sample. Blueprint-weighted selection instead calculates a target count for each domain. It takes the whole-number part first, then assigns the remaining questions to the largest fractional remainders. Ties follow the domain order in the table.

The ideal count for domain d is the round length multiplied by that domain's percentage weight.

q_{d} = \frac{N \times w_{d}}{100}

Here, N is the positive number of questions, w_d is 44, 28, 16 or 12, and q_d is the ideal domain count before whole-question rounding. For 20 questions, the ideal counts are 8.8, 5.6, 3.2 and 2.4. Rounding down allocates 18 questions; the two remaining places go to Fundamentals and Orchestration, producing 9, 6, 3 and 2.

Whole-question domain quotas for each supported mixed blueprint-weighted round size
Round questions	Fundamentals	Orchestration	Application delivery	Architecture
1	1	0	0	0
5	2	1	1	1
10	4	3	2	1
20	9	6	3	2
40	18	11	6	5
60	26	17	10	7
75	33	21	12	9

Weighted questions are selected without replacement within each eligible domain pool, then shuffled. The seed changes question selection and ordering, not the quotas. There are 29 foundational and 46 applied questions in total, but each separately filtered weighted mode offers only 1, 5, 10 or 20 questions because its available domain mix is narrower. Explicit review subsets, including missed-answer retries, keep their exact question scope rather than rebuilding blueprint quotas. Keeping a ten-question count across mixed, foundational and applied pools does not guarantee the same questions or result. No selection mode guarantees that every objective appears in a short round.

Counting points, accuracy and progress

Each correct answer earns 1 point and each incorrect answer earns 0. There is no partial credit, negative marking or difficulty multiplier. With C correct answers, A answered questions and a positive N total questions, points equal C and possible points equal N. Scores range from 0 to N points, equivalent to 0% to 100% round accuracy. Accuracy measures correctness; progress measures completion.

\begin{array}{l} Round accuracy (%) = 100 \times \frac{C}{N} \\ Progress (%) = 100 \times \frac{A}{N} \end{array}

Unanswered questions remain in the round-accuracy denominator but are not counted as incorrect or added to missed-answer review. Before completion, incorrect answers number A minus C, not N minus C. Final review requires an answer to every question.

Completion conditions and practice review labels, not official passing thresholds
Condition	Practice result
Answered questions are fewer than total questions	The round is incomplete; final review is not available.
All questions answered, with at least one incorrect answer	Start with the missed answers
All questions answered correctly	Perfect round

Weighting answered domains

Domain weighting gives each domain its declared share of the overall result rather than giving every sampled question the same influence. For domain d, let C_d be correct answers and A_d be answered questions. A domain's contribution is its weight multiplied by the fraction answered correctly.

\begin{array}{l} a_{d} = 100 \times \frac{C_{d}}{A_{d}} \\ p_{d} = w_{d} \times \frac{C_{d}}{A_{d}} \\ W = \sum_{d = 1}^{4} p_{d} \end{array}

a_d is domain accuracy in percent, p_d is the contribution in percentage points, and W is overall blueprint-weighted practice accuracy. These domain ratios require A_d greater than zero. A domain with no answers has unavailable accuracy and contribution, not zero accuracy. The overall weighted value is available only for a completed blueprint-weighted round with an answer in every domain and no restriction to an explicit review subset. Focused rounds and missed-answer retries do not receive that aggregate.

For a completed ten-question weighted round with correct counts 4, 0, 0, 0 and answered counts 4, 3, 2, 1, raw accuracy is 4/10 = 40%, while weighted accuracy is 44%. Answered-domain coverage sums the weights of domains with at least one answer, regardless of whether that answer was correct.

Answered-domain coverage (%) = \sum_{d : A_{d} > 0} w_{d}

Question and point totals are integers. Raw round accuracy and displayed progress are unrounded. Weighted accuracy, domain accuracy and contribution values are displayed to one decimal place; their displayed rounded values should not replace the underlying counts in calculations. In a completed 10-question weighted round, each correct Fundamentals answer contributes 11 percentage points to the weighted result. The single Architecture answer contributes 12 when correct. A short round's domain percentages therefore need to be read alongside its question counts.

Privacy Notes:

Answers are scored in your browser, but the saved-attempt URL contains the replay seed, round information and responses in an encoded, unencrypted form. Sharing that URL shares the attempt. It can remain in browser history and be sent to the site when the page is reopened, so browser scoring is not a guarantee that the attempt never leaves your device.

No account, exam credentials or certification details are required. Keep passwords and personal identifiers out of Replay seed, and treat exported answer ledgers as records containing your responses. Some chart and document functions load third-party resources, and opened study links use external sites. Fully offline operation is not guaranteed.

Worked Examples:

Eight correct answers with a different weighted percentage

Choose Blueprint-weighted practice, Mixed practice and 10 questions. A completed round has 3 of 4 Fundamentals answers correct, 2 of 3 Orchestration answers correct, both Application Delivery answers correct and its Architecture answer correct. The headline is 8/10, equivalent to 80% unweighted accuracy. The domain contributions total 79.666… percentage points before rounding, giving Blueprint-weighted practice accuracy of 79.7% after rounding. The difference is expected, not a lost answer. Review the two incorrect explanations rather than treating either percentage as a passing threshold.

A perfect one-question round with most domains untested

Select a one-question Blueprint-weighted practice round. Its whole-question allocation contains one Fundamentals question and none from the other domains. A correct answer produces 1/1 and Perfect round, but Answered-domain coverage is only 44% and the overall weighted practice accuracy remains unavailable. A five-question weighted round samples all four domains, although it still cannot establish understanding of every objective.

References:

Kubernetes and Cloud Native Associate (KCNA), The Linux Foundation.
Important Instructions: Multiple Choice Exams, The Linux Foundation.
Resources Allowed: All LF Certification Programs, The Linux Foundation.
CNCF Cloud Native Definition v1.1, Cloud Native Computing Foundation, 26 February 2024.
Containers, Kubernetes.
Pods, Kubernetes.
Kubernetes Components, Kubernetes.
Objects In Kubernetes, Kubernetes.
Deployments, Kubernetes.
Service, Kubernetes.
Liveness, Readiness, and Startup Probes, Kubernetes.
Signals, OpenTelemetry.