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 setting inside an Elyra pipeline node definition configures the specific allocation of CPU and memory for that proc. 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 setting inside an Elyra pipeline node definition configures the specific allocation of CPU and memory fo…
- Which OpenShift object defines the network security parameters that isolate model serving pods from other pro…
- Which interface allows data scientists to mount an external S3 bucket data source without coding access keys…
- WHICH SUB-COMPONENT WITHIN MODELMESH DIRECTLY MANAGES THE INTELLIGENT ROUTING AND PACKING OF MACHINE LEARNING…
- Which command lets you view the custom images that are currently registered and active within the RHOAI dashb…
- What is the result of applying a strict network policy that blocks ingress traffic to a data science project…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. What setting inside an Elyra pipeline node definition configures the specific allocation of CPU and memory for that processing step?
Answer: D. The runtime resource configurations options within the individual node properties menu
You configure CPU and memory allocations directly within the individual node properties menu using runtime resource configurations. Global workbench limits or machine profiles do not apply to specific pipeline execution steps.
Q2. Which OpenShift object defines the network security parameters that isolate model serving pods from other project namespaces?
Answer: D. NetworkPolicy custom resource configurations manifests
NetworkPolicy custom resources define strict network isolation rules to restrict pod traffic at the namespace level. Service Mesh handles advanced routing but NetworkPolicy provides the baseline Kubernetes security perimeter.
Q3. Which interface allows data scientists to mount an external S3 bucket data source without coding access keys directly?
Answer: C. OpenShift AI Data Connection UI
The OpenShift AI Data Connection UI allows users to securely mount external S3 buckets by creating Kubernetes secrets, abstracting away raw credentials. This provides seamless storage integration for workbenches without exposing keys.
Q4. WHICH SUB-COMPONENT WITHIN MODELMESH DIRECTLY MANAGES THE INTELLIGENT ROUTING AND PACKING OF MACHINE LEARNING MODELS?
Answer: D. proxy container inside runtime pod
The proxy container inside the runtime pod directly handles ModelMesh routing and packing. The Istio sidecar manages general network traffic, but it does not handle the internal model lifecycle or placement.
Q5. Which command lets you view the custom images that are currently registered and active within the RHOAI dashboard interface?
Answer: C. oc get imagecustomizations.dashboard.opendatahub.io -n redhat-ods-applications
Querying the imagecustomizations custom resource lists the active dashboard images. The other options use fabricated commands or query unrelated cluster configurations.
Q6. What is the result of applying a strict network policy that blocks ingress traffic to a data science project namespace?
Answer: C. External applications cannot query models deployed within that project namespace until explicit ingress exceptions are added
Blocking ingress traffic prevents external applications from querying deployed models until exceptions are added. Internal workbench coding remains functional because it relies on internal cluster networking.
Q7. WHAT COMPONENT OF OPENSHIFT AI LOGS THE HISTORICAL RECORDINGS OF ARTIFACTS PRODUCED BY PIPELINE EXPERIMENTS?
Answer: B. ML Metadata (MLMD) storage service layer
The ML Metadata service logs historical artifacts produced by pipeline experiments. Prometheus tracks cluster metrics, not specific pipeline artifact lineage or execution history.
Q8. Which container registry configuration must be adjusted if a custom notebook image requires an enterprise private registry credential?
Answer: B. Image Pull Secrets configuration inside the targeted workbench namespace or service account
Configuring Image Pull Secrets in the targeted namespace allows workbenches to authenticate to private registries. Git tokens handle repository access, while network routing is unrelated to image downloads.
Q9. Which serving architecture is best suited for deploying hundreds of small, disparate Scikit-Learn models on minimal infrastructure?
Answer: D. ModelMesh Framework Architecture
ModelMesh is explicitly designed for high-density deployments of multiple small models. KServe Serverless is better suited for scaling large, independent models rather than packing many disparate models together.
Q10. Which component manages the visual workspace environment where pipelines are developed visually inside JupyterLab?
Answer: D. Elyra Pipeline Editor extension installation bundles
The Elyra Pipeline Editor extension provides the visual workspace for building pipelines inside JupyterLab. While Tekton executes the pipelines underneath, Elyra is the actual visual interface component.
Q11. Which command displays the detailed real-time reconciliation logs of the KServe controller manager container?
Answer: C. oc logs -n redhat-ods-applications deployment/kserve-controller-manager
Querying the specific kserve-controller-manager deployment with the oc logs command reveals reconciliation events. The main operator deployment logs only show high-level installations, not specific controller actions.
Q12. Which custom resource definition is used to initialize cluster-wide basic prerequisites for the RHOAI platform operator?
Answer: D. DSCInitialization
The DSCInitialization custom resource is applied to configure cluster-wide prerequisites before components are deployed. Remember that DataScienceCluster manages individual components, while DSCInitialization handles the foundational setup for the operator.
Q13. Which interface can be used to upload an existing pipeline definition file if a user does not want to use the Python SDK?
Answer: D. Red Hat OpenShift AI Dashboard Pipelines Upload UI panel
The Red Hat OpenShift AI Dashboard provides a graphical upload panel for importing pipeline definition files without using the Python SDK. Look for the upload button directly within the Pipelines section.
Q14. Which component evaluates whether an incoming request to an InferenceService should scale up the existing deployment pod counts?
Answer: D. Knative Pod Autoscaler (KPA) or Horizontal Pod Autoscaler (HPA)
Knative relies on the Knative Pod Autoscaler to evaluate incoming request metrics and scale InferenceService deployments. Tekton and Advanced Cluster Management handle pipeline executions and multi-cluster governance, but they do not process real-time HTTP traffic scaling.
Q15. What occurs when an administrator changes a component's state from "Managed" to "Removed" inside the DataScienceCluster CR?
Answer: A. The operator automatically deletes all cluster resources associated with that specific component
Changing a component to Removed in the DataScienceCluster CR triggers the operator to delete all associated resources. The operator manages the entire lifecycle, so it will not merely pause management or leave orphaned pods running indefinitely in the namespace.
Q16. Which command configures a local git repository to use an enterprise internal proxy server within a workbench terminal?
Answer: B. git config –global http.proxy example.com
Git relies on standard git config commands to route its traffic through a proxy. On the exam, remember that workbenches behave as standard Linux pods, so standard Git CLI configurations apply rather than cluster-wide network commands.
Q17. What is the consequence of deleting the active Pipeline Server instance from a Data Science Project namespace?
Answer: D. All pipeline definitions, histories, and execution metadata records within that project are deleted
Deleting a Pipeline Server removes its underlying database, wiping all associated pipeline runs, definitions, and execution metadata. Workbenches and deployed models operate independently of the pipeline engine, so they remain completely unaffected.
Q18. What occurs when an InferenceService resource configuration changes from an OpenVINO runtime to a Triton runtime?
Answer: D. KServe triggers a rolling update deployment, spinning up the new runtime before terminating the old one
Changing the runtime in an InferenceService specification triggers a standard Kubernetes rolling update. KServe spins up the new runtime pods before terminating the old ones, ensuring zero downtime for model deployments.
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.