Practice for the Red Hat Certified Specialist in OpenShift AI (EX267) exam with 18 exam-style practice questions, instant answer reveals, and concise explanations of every correct answer. Topics include: WHAT BASELINE API CRD DEPLOYS PREDICTIVE ENDPOINTS PROVIDING AUTO-SCALING AND HEALTH CHECKS SPECIFIC TO INDEPENDENT MODE. Follow @CertPunch and visit certpunch.com for more certification practice exams and study content.
Prefer hands-on? Take this round as an interactive practice test — answer every question, get instant feedback, and see your score: Start the Red Hat Certified Specialist in OpenShift AI (EX267) practice test →
What you will practice
- WHAT BASELINE API CRD DEPLOYS PREDICTIVE ENDPOINTS PROVIDING AUTO-SCALING AND HEALTH CHECKS SPECIFIC TO INDEP…
- Which component ensures that user file changes inside a Jupyter notebook are preserved across a complete cont…
- What occurs behind the scenes when a data scientist selects "Clear volume contents" during workbench storage…
- Which CRD governs the operational lifecycle and feature flags of the OpenShift AI Dashboard interface?
- What is the primary role of the cert-manager operator when implementing KServe with OpenShift Service Mesh?
- What parameter in a KServe runtime configuration defines the port number that listens for incoming HTTP infer…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. WHAT BASELINE API CRD DEPLOYS PREDICTIVE ENDPOINTS PROVIDING AUTO-SCALING AND HEALTH CHECKS SPECIFIC TO INDEPENDENT MODELS?
Answer: B. InferenceService custom resource definition
The InferenceService custom resource deploys predictive endpoints with native auto-scaling and health checks. ServingRuntime defines the environment, but InferenceService manages the actual endpoint routing.
Q2. Which component ensures that user file changes inside a Jupyter notebook are preserved across a complete container pod replacement cycle?
Answer: B. Persistent Volume Claims (PVC) mapped to the notebook directory structure
Persistent Volume Claims mapped to the notebook directory retain user files across pod restarts. Ephemeral container storage is wiped upon pod replacement, making PVCs essential for data science workbench persistence.
Q3. What occurs behind the scenes when a data scientist selects "Clear volume contents" during workbench storage deletion operations?
Answer: C. The persistent volume claim is deleted, triggering the storage plugin to destroy the data disk
Selecting this option deletes the persistent volume claim, which tells the storage plugin to destroy the backing disk. OpenShift AI does not use a trash folder, so clearing the volume permanently deletes the files.
Q4. Which CRD governs the operational lifecycle and feature flags of the OpenShift AI Dashboard interface?
Answer: C. OdhDashboardConfig
The correct answer works because the OdhDashboardConfig custom resource manages lifecycle settings and feature flags for the dashboard. The distractors use plausible but ultimately fake names, so remembering the specific OpenDataHub prefix is crucial.
Q5. What is the primary role of the cert-manager operator when implementing KServe with OpenShift Service Mesh?
Answer: C. Generating and managing TLS certificates for secure service-to-service communication
The correct answer works because cert-manager generates and manages the TLS certificates required for secure mTLS communication between Service Mesh components. Other options confuse certificate management with idle notebook encryption or cluster auto-scaling operations.
Q6. What parameter in a KServe runtime configuration defines the port number that listens for incoming HTTP inference queries?
Answer: A. containerPort configuration parameter inside the serving runtime container spec
The containerPort parameter within the serving runtime container spec defines the network port listening for incoming HTTP inference queries. External routes merely forward traffic to this underlying container port.
Q7. What happens if the requested CPU limits in a workbench size configuration exceed the remaining unallocated CPU resources of the cluster nodes?
Answer: B. The workbench pod remains Pending due to insufficient resource capacities
When a workbench requests CPU limits exceeding available node capacity, the scheduler leaves the pod in a Pending state due to insufficient resources. OpenShift will never silently lower requested container limits to force scheduling.
Q8. How can an administrator completely reset a specific user's JupyterLab workbench profile data without affecting other users?
Answer: C. Delete the individual Persistent Volume Claim associated with that user workbench
Deleting the individual Persistent Volume Claim associated with a workbench completely resets that specific user's profile data. Restarting the notebook controller will not delete user storage, and reinstalling the operator affects all users.
Q9. How can you determine the specific version of the Red Hat OpenShift AI platform currently installed on a cluster?
Answer: B. View the CSV details within the Installed Operators section of the OpenShift console
Checking the ClusterServiceVersion within the Installed Operators tab shows the exact OpenShift AI version. This represents the standard Operator Lifecycle Manager method for verifying installed applications and their respective versions.
Q10. What is the main role of the KServe container named queue-proxy placed alongside the model predictor container?
Answer: A. To intercept incoming traffic, manage request concurrency, and collect metrics data for autoscaling
The queue-proxy container intercepts incoming traffic, monitors concurrency, and exposes metrics. KServe uses these metrics to scale the predictor accurately, separate from the storage and runtime tasks.
Q11. Which network protocol is utilized internally by ModelMesh to communicate between the proxy and runtime containers?
Answer: D. gRPC
ModelMesh relies on gRPC for high-performance internal communication between the proxy and runtime containers. This protocol efficiently handles the high-volume, low-latency routing required for scalable inference.
Q12. What parameter in a Data Connection defines the exact address used to reach a private on-premise S3-compatible storage cluster?
Answer: A. Endpoint URL parameter setting
The Endpoint URL parameter in a Data Connection specifies the exact address used to reach a private S3-compatible storage cluster. Region and bucket parameters help structure data but do not define the network endpoint itself.
Q13. What happens if you attempt to launch a workbench with a persistent volume claim that is already mounted by an active workbench?
Answer: D. The new workbench pod fails to schedule due to a Multi-Attach error
A persistent volume claim already mounted by an active workbench prevents a new workbench from scheduling due to a Multi-Attach error. Read-only mounting or automatic data cloning are not default OpenShift storage behaviors.
Q14. Which resource controls memory constraints assigned to data science workbench types so users do not overallocate cluster platforms?
Answer: C. Notebook Size profiles configured via Dashboard settings configs
Notebook Size profiles define the CPU and memory limits presented to users when launching workbenches. This prevents users from selecting container sizes that exceed cluster capacity or consuming all available resources.
Q15. What parameter in a KServe InferenceService resource enables the configuration of canary deployments for a new model version?
Answer: C. traffic percent routing configuration values parameters
The traffic percent parameter controls how incoming requests are split between different model versions. Adjusting these values enables seamless canary rollouts without altering the underlying storageUri or memoryLimit settings.
Q16. What configuration step seamlessly binds an associated S3 data source so an experimental notebook reads target images directly?
Answer: D. Creating a Data Connection inside the Data Science Project via OpenShift AI Dashboard UI
Creating a Data Connection securely binds S3 credentials to the project environment variables. Workbenches automatically consume this connection without requiring manual text files or exposing raw secrets in the notebook directory.
Q17. What is the function of the ClusterServingRuntime custom resource in Red Hat OpenShift AI?
Answer: B. Defining reusable serving container templates for different model frameworks
The ClusterServingRuntime custom resource defines reusable serving container templates for various model frameworks. It configures the deployment environment for models, leaving CPU allocation and network routing to standard OpenShift mechanisms.
Q18. Which object handles the conversion of internal cluster services into accessible public URLs for applications outside the cluster?
Answer: A. Route or Ingress custom resource objects definitions
Routes expose internal Kubernetes services to external traffic by mapping public URLs to service ports. Services alone only provide internal cluster networking, while ConfigMaps and NetworkPolicies handle configuration and traffic restrictions, respectively.
More Red Hat Certified Specialist in OpenShift AI (EX267) drills and other practice exams are on @CertPunch. New rounds drop every few days at certpunch.com.