Practice for the Red Hat Certified Specialist in OpenShift AI (EX267) exam with 17 exam-style practice questions, instant answer reveals, and concise explanations of every correct answer. Topics include: Which element is required to authenticate a workbench session to an external secure enterprise Git repository?. 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
- Which element is required to authenticate a workbench session to an external secure enterprise Git repository?
- How can data scientists share custom notebook images cluster-wide within OpenShift AI?
- How can a data scientist update the code implementation of a pipeline step without rewriting the entire pipel…
- How can an administrator enforce strict memory boundaries on a developer's Data Science Project namespace to…
- What component handles monitoring and alerts metrics aggregation for OpenShift AI clusters?
- What is the default behavior of KServe when an application experiences a massive spike in concurrent model in…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. Which element is required to authenticate a workbench session to an external secure enterprise Git repository?
Answer: A. A personal access token or SSH key configured in the Jupyter environment or Git credential helper
Workbenches authenticate securely to external Git repositories using configured credentials like an SSH key or a personal access token within the environment. Expect to configure this routinely when cloning private model training code.
Q2. How can data scientists share custom notebook images cluster-wide within OpenShift AI?
Answer: A. Register the image in the OpenShift AI dashboard settings
Custom notebook images are registered cluster-wide in the OpenShift AI dashboard settings under notebook environments. Pushing to a namespace does not register it, and editing the operator CRs is not the administrative workflow for adding custom images.
Q3. How can a data scientist update the code implementation of a pipeline step without rewriting the entire pipeline visual graph?
Answer: A. Update the source code file or notebook cell referenced by that specific node and recompile/run the pipeline
Pipeline nodes execute specific files or cells, so updating the underlying code and rerunning the pipeline is correct. Modifying cluster ingress or reinstalling the operator does not update the logic within a pipeline step.
Q4. How can an administrator enforce strict memory boundaries on a developer's Data Science Project namespace to avoid cluster resource exhaustion?
Answer: C. By creating a ResourceQuota object inside that specific project namespace
OpenShift uses ResourceQuota objects to enforce strict memory and CPU limits within a namespace. Modifying container images or writing cron scripts to delete pods are not standard administrative practices for managing resource boundaries.
Q5. What component handles monitoring and alerts metrics aggregation for OpenShift AI clusters?
Answer: D. Prometheus
OpenShift uses Prometheus to aggregate monitoring metrics and trigger alerts for cluster components. Jaeger handles tracing, while Fluentd and ElasticSearch are typically used for log collection and search.
Q6. What is the default behavior of KServe when an application experiences a massive spike in concurrent model inference requests?
Answer: C. It scales up the number of replica pods automatically based on concurrency or request metrics
KServe uses Knative Serving to automatically scale replica pods based on configured concurrency targets. Crashing pods or returning 500 errors is not the standard behavior for scaling under heavy inference load.
Q7. Which configuration inside an Elyra pipeline specifies the exact container image a specific pipeline node should execute within?
Answer: D. Runtime Image setting
The Elyra Runtime Image property dictates the exact container environment used for a specific node execution. Environment variables are for passing data, while node limits only restrict hardware consumption.
Q8. To ensure that an OpenShift AI workload only runs on nodes equipped with NVIDIA GPUs, what Kubernetes mechanism is applied to the pod?
Answer: D. Taints and Tolerations along with Node Selectors
Taints and tolerations ensure pods are permitted to run on specialized GPU nodes. Network policies and quotas manage traffic and resource limits, respectively, and do not dictate specific node hardware placement.
Q9. What configuration item maps an active Git repository directly into a Jupyter workbench environment?
Answer: A. Git Repository URL in workbench creation settings
Specifying a Git Repository URL during workbench creation clones the repository directly into the environment. ConfigMaps or emptyDir volumes do not automatically sync external version control history.
Q10. Which OpenShift resource is created when an OpenShift AI pipeline run is executed?
Answer: C. Tekton PipelineRun
OpenShift AI executes pipeline runs using Tekton, creating a Tekton PipelineRun resource. Argo Workflows or Kubernetes Jobs are not the underlying resources used by OpenShift AI pipelines.
Q11. What must be configured on an OpenShift cluster to ensure that container images are pulled successfully from a private corporate Quay.io registry?
Answer: B. A global or namespace-specific Image Pull Secret
An Image Pull Secret provides the necessary authentication to pull images from a private registry. OpenShift requires these secrets to authorize access to private Quay.io repositories.
Q12. Which command-line utility is primary for inspecting OpenShift cluster resources during troubleshooting?
Answer: A. oc
The oc command-line utility is the primary tool for inspecting and managing OpenShift resources. Dedicated CLI tools like rhoai-cli or odhctl are not standard requirements.
Q13. What is the consequence of configuring a workbench to use a Container Image that lacks a Python installation?
Answer: B. The JupyterLab interface fails to start or cannot spawn notebook kernels
JupyterLab requires a Python installation to spawn notebook kernels successfully. Without Python, the workbench fails to operate as an interactive data science environment.
Q14. What status indicator confirms that the OpenShift AI console deployment is complete?
Answer: C. Ready state in the DataScienceCluster Custom Resource
The DataScienceCluster custom resource reaching a Ready state confirms the operator finished reconciling and the OpenShift AI components are deployed. Pod statuses alone do not validate installation success, because dependent controllers might still be initializing.
Q15. Where does an administrator find the precise logs for troubleshooting a failure during the installation phase of the OpenShift AI components?
Answer: D. The operator pod logs in the redhat-ods-operator namespace
The operator pod logs in the redhat-ods-operator namespace capture the installation and reconciliation steps for OpenShift AI components. Workbench container traces appear only after the workbench is successfully created, which happens too late for troubleshooting initial platform failures.
Q16. When configuring Multi-Model serving platforms in OpenShift AI, which runtime server is selected by default for OpenVINO models?
Answer: D. ModelMesh OpenVINO Model Server
OpenVINO Model Server is the default runtime integrated with ModelMesh to optimize and serve models on OpenShift AI. Other runtimes like Triton or KServe are available for multi-framework support but are not the default selection for OpenVINO workflows.
Q17. What authentication mechanism does the OpenShift AI dashboard use by default to verify a user's identity?
Answer: D. OpenShift Container Platform OAuth service
OpenShift AI delegates user authentication directly to the underlying OpenShift Container Platform OAuth service. This integration means identity provider configuration is managed centrally on the cluster itself rather than duplicated inside OpenShift AI.
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.