Practice Exams:

Linux Foundation KCNA Dumps

EXAM: Kubernetes and Cloud Native Associate
Verified By Experts
Premium Bundle
$40.00
$59.97 $99.97

KCNA Premium Bundle


  • Premium File 333 Questions & Answers
  • Training Course 54 Lectures
  • Study Guide 410 Pages

Linux Foundation KCNA Practice Test 35

Question 340

A factory sensor aggregator runs slowly under CPU throttling despite successful Pod placement. Which engineering change and test jointly addresses the scenario?

  1. Route a measured fraction of traffic to a canary or switch between separately verified blue and green environments using an appropriate routing controller. Acceptance evidence: In a staging gateway route 5 percent to a canary, inspect version-specific success rates and verify a rollback rule.
  2. Set realistic CPU and memory requests for Pods and investigate FailedScheduling when no node has sufficient allocatable resources. Acceptance evidence: Set an intentionally excessive memory request in a lab and inspect Pod Pending Events before adjusting it.
  3. Review requests, limits, QoS class and termination reasons rather than interpreting every restart as a scheduler failure. Acceptance evidence: Create a synthetic CPU-intensive workload with tight CPU limits and compare throttling metrics with an OOMKilled test Pod.
  4. Choose Pod affinity only when co-locating related workloads is beneficial and verify independent network policies still govern traffic. Acceptance evidence: Apply a preferred Pod affinity against a cache Pod label and inspect node placement while testing access policy separately.

Explanation

Correct Answer

C:

The situation is that a factory sensor aggregator runs slowly under CPU throttling despite successful Pod placement. Review requests, limits, QoS class and termination reasons rather than interpreting every restart as a scheduler failure. Acceptance evidence: Create a synthetic CPU-intensive workload with tight CPU limits and compare throttling metrics with an OOMKilled test Pod. This pairing is justified because CPU limits may throttle running processes, whereas exceeding memory constraints can cause OOM termination; Pod scheduling itself depends on requests.

Incorrect Answers

A:

This action and check address the distinct objective “Distinguish canary and blue-green delivery strategies from Kubernetes Deployment defaults” instead of the stated acceptance requirement. The matching action is to review requests, limits, QoS class and termination reasons rather than interpreting every restart as a scheduler failure. Its confirming test is to create a synthetic CPU-intensive workload with tight CPU limits and compare throttling metrics with an OOMKilled test Pod.

B:

This action and check address the distinct objective “Use container resource requests to help the scheduler fit Pods on nodes” instead of the stated acceptance requirement. The matching action is to review requests, limits, QoS class and termination reasons rather than interpreting every restart as a scheduler failure. Its confirming test is to create a synthetic CPU-intensive workload with tight CPU limits and compare throttling metrics with an OOMKilled test Pod.

D:

This action and check address the distinct objective “Use Pod affinity for co-location without assuming it creates networking permissions” instead of the stated acceptance requirement. The matching action is to review requests, limits, QoS class and termination reasons rather than interpreting every restart as a scheduler failure. Its confirming test is to create a synthetic CPU-intensive workload with tight CPU limits and compare throttling metrics with an OOMKilled test Pod.

Question 341

A research cluster needs one log-collection agent on each worker. What technical action should the platform team take in this situation?

  1. Use cluster DNS records such as service.namespace.svc.cluster.local when a Pod connects to Services in other namespaces.
  2. Choose Pod affinity only when co-locating related workloads is beneficial and verify independent network policies still govern traffic.
  3. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance.
  4. Inspect kubelet-reported Pod and container conditions when a scheduled workload fails to start, while keeping cluster resources managed through the API.

Explanation

Correct Answer

C:

Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. This resolves the issue because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count. To verify, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Incorrect Answers

A:

This option addresses the distinct objective “Verify CoreDNS resolution and namespace-qualified Service names”, not the immediate requirement. CoreDNS watches Service objects and serves cluster-aware names; bare names normally resolve relative to the requesting Pod's namespace and search domains. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

B:

This option addresses the distinct objective “Use Pod affinity for co-location without assuming it creates networking permissions”, not the immediate requirement. Pod affinity affects scheduler placement relative to labeled Pods in selected topology domains, not application authentication or firewall rules. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

D:

This option addresses the distinct objective “Recognize the kubelet role in local Pod lifecycle management”, not the immediate requirement. Each kubelet watches assigned Pod specifications and coordinates the container runtime, volumes and health information on its node. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Question 342

A research cluster needs one log-collection agent on each worker. What underlying Kubernetes behavior is most relevant to this technical decision?

  1. Service type determines how clients can reach a stable Service: ClusterIP is internal, NodePort exposes a port on nodes, and LoadBalancer integrates external infrastructure.
  2. Readiness affects endpoint eligibility while liveness can cause restarts; a startup probe defers other probe checks until initialization succeeds.
  3. A DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.
  4. Topology spread constraints compare Pod distribution in topology domains and can restrict or prefer placement depending on whenUnsatisfiable.

Explanation

Correct Answer

C:

A DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count. It supports this action: Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. To confirm, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Incorrect Answers

A:

That statement explains the distinct objective “Choose ClusterIP, NodePort and LoadBalancer Service types for appropriate reachability” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

B:

That statement explains the distinct objective “Understand readiness, liveness and startup probes as different signals” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

D:

That statement explains the distinct objective “Spread replicas across failure domains with topologySpreadConstraints” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Question 343

A research cluster needs one log-collection agent on each worker. What acceptance check most directly confirms this requirement is met?

  1. Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.
  2. Apply a preferred Pod affinity against a cache Pod label and inspect node placement while testing access policy separately.
  3. Verify a synthetic exporter is listed UP in Prometheus target status and query its request counter after generating traffic.
  4. Observe a controlled drain with replica owners and a PDB, then verify workloads move and uncordon restores eligibility.

Explanation

Correct Answer

A:

Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment. This is the relevant test because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

B:

This check would help evaluate the distinct objective “Use Pod affinity for co-location without assuming it creates networking permissions”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

C:

This check would help evaluate the distinct objective “Understand Prometheus pull-based metrics collection and exporters”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

D:

This check would help evaluate the distinct objective “Use cordon and drain with disruption considerations for planned node maintenance”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Question 344

A research cluster needs one log-collection agent on each worker. The team suggests: “Use a Deployment with replicas=3 and assume it guarantees exactly one collector on each of three worker nodes.” Which failure should the reviewer anticipate?

  1. The scheduler may legally use a non-preferred node, violating the policy that was mistakenly assumed mandatory.
  2. The scheduler could place multiple replicas together, leaving a node without the intended local collector.
  3. Cross-service latency remains difficult to diagnose because spans cannot be tied to one request path.
  4. Compromised credentials can perform widespread cluster changes without attribution or least privilege.

Explanation

Correct Answer

B:

The scheduler could place multiple replicas together, leaving a node without the intended local collector. The shortcut creates this risk because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

A:

That outcome concerns the distinct objective “Distinguish preferred node affinity from required scheduling rules”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

C:

That outcome concerns the distinct objective “Instrument services with OpenTelemetry context propagation”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

D:

That outcome concerns the distinct objective “Use least-privilege kubectl authentication and authorization during routine administration”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

Question 345

A research cluster needs one log-collection agent on each worker. Which engineering change and test jointly addresses the scenario?

  1. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Acceptance evidence: Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.
  2. Define a user-centered service level indicator and objective, such as percentage of successful requests within a latency threshold over a stated window. Acceptance evidence: Calculate a synthetic seven-day success ratio from request metrics and compare it to the agreed target before a risky rollout.
  3. Pin validated production images by registry digest and promote the same signed artifact across environments. Acceptance evidence: Resolve an approved image digest, deploy that reference in staging and verify the same digest exists in the production Pod spec.
  4. Apply PriorityClasses sparingly for essential workloads and plan capacity so preemption does not unnecessarily evict useful lower-priority Pods. Acceptance evidence: Simulate controlled resource pressure with two priorities and inspect Events for whether a high-priority test Pod can schedule.

Explanation

Correct Answer

A:

Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Acceptance evidence: Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment. This pairing is justified because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

B:

This action and check address the distinct objective “Use SLI and SLO measures to focus reliability objectives” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

C:

This action and check address the distinct objective “Use an immutable image digest for exact deployment provenance” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

D:

This action and check address the distinct objective “Use PriorityClasses and understand preemption without assuming they guarantee availability” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Question 346

A retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. Which engineering approach is the most appropriate for this cloud-native case?

  1. Use requiredDuringSchedulingIgnoredDuringExecution for mandatory constraints and preferredDuringSchedulingIgnoredDuringExecution for a soft placement preference.
  2. Apply ingress and egress NetworkPolicy to selected Pods and verify both permitted and denied synthetic flows under a network plugin that implements policy.
  3. Separate compile and runtime stages to retain only required binary and runtime dependencies in the deployed container image.
  4. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance.

Explanation

Correct Answer

D:

The situation is that a retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. This resolves the issue because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count. To verify, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Incorrect Answers

A:

This option addresses the distinct objective “Distinguish preferred node affinity from required scheduling rules”, not the immediate requirement. A required affinity expression blocks scheduling onto nonmatching nodes; a preferred expression influences scoring but permits fallback when necessary. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

B:

This option addresses the distinct objective “Recognize NetworkPolicy isolation only when the CNI enforces it”, not the immediate requirement. NetworkPolicy declares traffic restrictions, but Kubernetes itself does not provide enforcement when the installed CNI lacks NetworkPolicy support. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

C:

This option addresses the distinct objective “Build smaller images with multistage builds and remove unnecessary tooling”, not the immediate requirement. Multistage builds reduce image contents and potential attack surface by keeping build toolchains and source artifacts out of the final runtime layer. For the stated case, a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Question 347

A retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. What infrastructure principle makes the correct treatment effective?

  1. Prometheus commonly pulls numeric samples from configured targets, stores them as labeled time series and supports PromQL queries over those samples.
  2. Under pressure, a higher-priority pending Pod may preempt eligible lower-priority Pods, but constraints and eviction rules still apply.
  3. A supported runtime such as containerd executes container operations requested through kubelet using the Container Runtime Interface.
  4. A DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Explanation

Correct Answer

D:

The situation is that a retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. A DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count. It supports this action: Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. To confirm, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Incorrect Answers

A:

That statement explains the distinct objective “Understand Prometheus pull-based metrics collection and exporters” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

B:

That statement explains the distinct objective “Use PriorityClasses and understand preemption without assuming they guarantee availability” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

C:

That statement explains the distinct objective “Explain the container runtime interface without conflating it with the control plane” rather than the cause in this case. The deciding mechanism here is that a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Question 348

A retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. What verification result would most convincingly distinguish a working solution from a claim?

  1. Inspect a failed Pod's last terminated reason and check whether an unintended command override prevented the server from starting.
  2. Deploy a synthetic stateless workload to two conformant clusters and record which storage, ingress and identity settings require changes.
  3. Set an intentionally excessive memory request in a lab and inspect Pod Pending Events before adjusting it.
  4. Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Explanation

Correct Answer

D:

The situation is that a retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment. This is the relevant test because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

A:

This check would help evaluate the distinct objective “Understand ENTRYPOINT and command overrides when a container exits immediately”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

B:

This check would help evaluate the distinct objective “Apply cloud-native portability and immutable infrastructure principles pragmatically”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

C:

This check would help evaluate the distinct objective “Use container resource requests to help the scheduler fit Pods on nodes”, but not the required result. For this case, add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Question 349

A retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. The change review contains this proposal: “Use a Deployment with replicas=3 and assume it guarantees exactly one collector on each of three worker nodes.” What risk matters most?

  1. The scheduler could place multiple replicas together, leaving a node without the intended local collector.
  2. The Service still has no ready endpoints and clients continue to encounter unavailable responses.
  3. The Pod remains unschedulable because requested capacity and node feasibility did not change.
  4. A scheduled workload may violate performance expectations because the declared label was not backed by infrastructure inventory.

Explanation

Correct Answer

A:

The situation is that a retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. The scheduler could place multiple replicas together, leaving a node without the intended local collector. The shortcut creates this risk because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

B:

That outcome concerns the distinct objective “Investigate readiness failures separately from liveness-related restarts”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

C:

That outcome concerns the distinct objective “Use container resource requests to help the scheduler fit Pods on nodes”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

D:

That outcome concerns the distinct objective “Use nodeSelector to place a workload on nodes with required labels”; it is not the most direct consequence of this shortcut. The relevant consequence is that the scheduler could place multiple replicas together, leaving a node without the intended local collector.

Question 350

A retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. Which implementation together with matching acceptance evidence best meets this case?

  1. Read a workload's spec and status separately and inspect conditions when determining whether an API change became healthy. Acceptance evidence: Apply a test manifest with impossible node constraints, then inspect the accepted spec and Pending status/Events.
  2. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Acceptance evidence: Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.
  3. Give clients a Service DNS name rather than hardcoding Pod IP addresses that can change when replicas are replaced. Acceptance evidence: Delete and replace a Deployment Pod and verify a client continues to reach its Service DNS name despite changed backend Pod addresses.
  4. Choose Pod affinity only when co-locating related workloads is beneficial and verify independent network policies still govern traffic. Acceptance evidence: Apply a preferred Pod affinity against a cache Pod label and inspect node placement while testing access policy separately.

Explanation

Correct Answer

B:

The situation is that a retail company adds branch Kubernetes nodes and wants a node-local metrics exporter automatically installed. Run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Acceptance evidence: Add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment. This pairing is justified because a DaemonSet controller targets eligible nodes for a Pod each, adapting to node additions and labels rather than maintaining a global arbitrary replica count.

Incorrect Answers

A:

This action and check address the distinct objective “Interpret Kubernetes spec and status fields without confusing intent with condition” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

C:

This action and check address the distinct objective “Distinguish Pod IPs from stable Services during rescheduling” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

D:

This action and check address the distinct objective “Use Pod affinity for co-location without assuming it creates networking permissions” instead of the stated acceptance requirement. The matching action is to run a logging or node monitoring agent as a DaemonSet when every eligible worker needs its own instance. Its confirming test is to add a test worker node and inspect the corresponding DaemonSet agent Pod without scaling a Deployment.

Kubernetes and Cloud Native Associate (KCNA) Practice Tests