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 is the standard tool used to convert a standard Python function into an executable component inside an OpenShift AI. 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 is the standard tool used to convert a standard Python function into an executable component inside an O…
- What occurs when an administrator changes a component state to Removed within the DataScienceCluster configur…
- Which component ensures that a data science project namespace contains the necessary network policies for iso…
- What is the function of the Caikit component when used within the Caikit+TGIS model serving runtime framework?
- What is the purpose of the DSCInitialization Custom Resource in the Red Hat OpenShift AI operator architectur…
- What type of backend storage structure is created when configuring an internal database for the pipeline serv…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. What is the standard tool used to convert a standard Python function into an executable component inside an OpenShift AI pipeline?
Answer: A. kfp.components.create_component_from_func
The kfp.components.create_component_from_func function converts standard Python functions into executable pipeline components. The other options use fabricated module paths that do not exist in the Kubeflow Pipelines SDK.
Q2. What occurs when an administrator changes a component state to Removed within the DataScienceCluster configuration?
Answer: B. The operator automatically deletes the deployment assets associated with that component
Setting the component state to Removed triggers the OpenShift AI Operator to automatically delete the related deployment assets. On the exam, remember that toggling states in the DataScienceCluster directly provisions or tears down resources rather than just hiding them.
Q3. Which component ensures that a data science project namespace contains the necessary network policies for isolation?
Answer: D. Red Hat OpenShift AI Operator
The OpenShift AI Operator configures the necessary network policies to isolate namespaces when a Data Science Project is created. Standard Kubernetes controllers handle routing and events, but they do not inject project-specific isolation rules automatically.
Q4. What is the function of the Caikit component when used within the Caikit+TGIS model serving runtime framework?
Answer: A. To provide a developer-friendly API layer and abstraction wrapper for model inference requests
Caikit acts as a developer-friendly API layer that wraps and standardizes inference requests before passing them to the underlying serving engine. TGIS handles the actual model inference and GPU memory allocation, not the API abstraction.
Q5. What is the purpose of the DSCInitialization Custom Resource in the Red Hat OpenShift AI operator architecture?
Answer: B. To configure core global prerequisites, proxy setups, and security tracking profiles for RHOAI
DSCInitialization sets global prerequisites like proxies and security profiles before deploying components. It operates as a cluster-wide initialization step, whereas the main DataScienceCluster resource enables individual components.
Q6. What type of backend storage structure is created when configuring an internal database for the pipeline server?
Answer: A. A Persistent Volume Claim backing a database container within the project
An internal pipeline server database uses a Persistent Volume Claim to provision storage for its database container within the project namespace. Choosing external cloud-managed endpoints would indicate you selected an external database configuration rather than an internal one.
Q7. What parameter determines the minimum number of model server pod instances KServe must keep alive even when zero traffic occurs?
Answer: A. minReplicas scaling configuration property values
The minReplicas parameter dictates exactly how many model server pod instances KServe must keep actively running regardless of incoming traffic. If minReplicas is set to zero, the deployment scales down completely, which eliminates zero-traffic operational overhead.
Q8. What is the primary indicator that an OpenShift cluster node has the NVIDIA GPU capabilities successfully configured and ready for workloads?
Answer: C. The node resource capacities include nvidia.com allocations numbers
A node is successfully configured for GPU workloads when its resource capacities explicitly list available allocations under the nvidia.com extended resource namespace. Simply having hardware attached does not guarantee readiness; the GPU Operator must advertise these compute allocations.
Q9. Which custom resource configuration maps a custom container image to the ModelMesh multi-model serving options?
Answer: A. ServingRuntime custom resource configuration layout
The ServingRuntime custom resource defines the runtime container image and configuration used by ModelMesh for multi-model serving. Workbench ImageStreams are completely unrelated to inference deployments and only provide notebook environments.
Q10. What happens to a Pipeline Run if the underlying Tekton pod is deleted mid-execution?
Answer: B. The pipeline status transitions to Failed
Deleting the underlying Tekton execution pod causes the active PipelineRun task to fail immediately, transitioning the overall run status to failed. Tekton does not automatically pause or retry the pipeline without explicit error handling configured.
Q11. Where are the logs of a model server container running under ModelMesh stored for standard administrator collection?
Answer: A. Within the individual model mesh serving runtime container pods in the project namespace
ModelMesh runs as containers inside standard Kubernetes pods, meaning administrators collect logs directly from the runtime pods in the serving namespace. Persistent object storage contains artifacts, not runtime logs.
Q12. What is the structural role of the ModelMesh controller container named mm-lines-proxy?
Answer: C. It handles internal orchestration routing and model life-cycle tasks within the multi-model pods
The proxy container handles internal routing and orchestration tasks within the multi-model pods. Storage and volume encryption responsibilities are handled by standard Kubernetes primitives, not this specific controller.
Q13. Which tool provides automated code delivery capabilities to maintain Red Hat OpenShift AI resources via GitOps workflows?
Answer: C. Red Hat OpenShift GitOps Operator (Argo CD framework)
The Red Hat OpenShift GitOps Operator uses Argo CD to declaratively manage cluster resources via Git repositories. While Tekton handles CI/CD pipeline execution, GitOps specifically ensures continuous state synchronization.
Q14. What custom resource system is generated to run an individual tracked workflow within the pipelines section?
Answer: B. PipelineRun
OpenShift AI pipelines use Tekton under the hood, meaning executing an entire pipeline creates a PipelineRun custom resource. TaskRun resources only execute individual pipeline steps rather than the entire workflow.
Q15. What controller watches and stabilizes intraclass certificate renewals ensuring infrastructure parts encrypt communications safely?
Answer: C. Cert-Manager Operator system components providing automated certificate provisioning workflows
The Cert-Manager Operator handles automated certificate provisioning and renewal workflows across the cluster. Service Mesh components primarily encrypt service-to-service traffic but rely on infrastructure certificates provisioned by the certificate manager.
Q16. Which CRD resource manifests are edited directly by admins to expand global inference compatibilities by creating custom model servers?
Answer: D. ServingRuntime / ClusterServingRuntime custom resources blueprints added into cluster environments
Administrators define ServingRuntime or ClusterServingRuntime custom resources to deploy custom model servers. The DataScienceCluster CR configures core operator components, but ServingRuntimes dictate specific inference deployment environments.
Q17. Which Red Hat Operator provides the multi-tenant network infrastructure required by KServe?
Answer: B. Red Hat OpenShift Service Mesh
KServe relies on Red Hat OpenShift Service Mesh for its multi-tenant networking and traffic routing capabilities. Serverless manages autoscaling, but it cannot provide the required network isolation alone.
Q18. What parameter inside a Data Connection configuration contains the secret authentication password token string for object storage?
Answer: D. AWS_SECRET_ACCESS_KEY
AWS_SECRET_ACCESS_KEY stores the secret password token required for object storage authentication. AWS_ACCESS_KEY_ID functions as the public username, while the endpoint and bucket parameters define the connection target.
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.