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: Which serving architecture is designed for advanced single-model serving including Large Language Models (LLMs) with adv. 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 serving architecture is designed for advanced single-model serving including Large Language Models (LLM…
- What parameter in the InferenceService configuration specifies the path to the model file within an object st…
- Which tool allows automated delivery and configuration of RHOAI projects across multiple clusters from a Git…
- Which command-line tool allows you to directly interact with OpenShift AI custom resources like DataScienceCl…
- What is the purpose of the DSCInitialization custom resource in Red Hat OpenShift AI?
- What role-based access control (RBAC) permission level is required to create a new Data Science Project withi…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. Which serving architecture is designed for advanced single-model serving including Large Language Models (LLMs) with advanced routing?
Answer: A. KServe
KServe is the correct choice because it provides advanced, single-model serving capabilities specifically optimized for large language models and complex routing. ModelMesh is the strongest distractor, but it targets high-density, multi-model serving rather than single-model routing.
Q2. What parameter in the InferenceService configuration specifies the path to the model file within an object storage bucket?
Answer: D. storageUri
The storageUri parameter correctly defines the exact path to the model files within an object storage bucket during inference configuration. Options like bucketPath fail because they are not valid parameters in the custom resource definition.
Q3. Which tool allows automated delivery and configuration of RHOAI projects across multiple clusters from a Git repository?
Answer: C. Red Hat OpenShift GitOps (Argo CD)
Red Hat OpenShift GitOps leverages Argo CD to declaratively automate application delivery and configuration across clusters from a Git repository. Elyra is a distractor that only handles visual pipeline execution within a notebook environment.
Q4. Which command-line tool allows you to directly interact with OpenShift AI custom resources like DataScienceClusters?
Answer: D. kubectl/oc
The kubectl and oc command-line tools are correct because they directly manage Kubernetes custom resources like DataScienceClusters. Helm is a package manager and does not interact with custom resources natively.
Q5. What is the purpose of the DSCInitialization custom resource in Red Hat OpenShift AI?
Answer: A. It configures the proxy and certificate settings for the operator
The DSCInitialization custom resource correctly applies global proxy and certificate settings required by the operator before components deploy. Launching Jupyter servers fails as a distractor because that is handled by workbench resources later.
Q6. What role-based access control (RBAC) permission level is required to create a new Data Science Project within OpenShift AI?
Answer: A. Dedicated cluster user or project administrator privileges
Creating a Data Science Project maps directly to creating an OpenShift namespace. You must have standard project creation privileges or cluster administrator rights. Read-only or storage management roles will prevent namespace creation.
Q7. What occurs when an administrator changes the management state of a component to Managed in the DataScienceCluster CR?
Answer: A. The operator actively reconciles, deploys, and updates that specific component to ensure health
Setting the management state to Managed tells the operator to actively deploy and maintain that component. If set to Removed, the operator deletes the component instead of updating it.
Q8. When creating a Data Science Project in OpenShift AI, what underlying OpenShift object is automatically provisioned?
Answer: A. Namespace
Creating a Data Science Project automatically provisions an underlying OpenShift Namespace to provide logical isolation. The namespace securely houses all related resources, such as workbenches, pipelines, and deployed models, for that specific project.
Q9. Which role must be bound to a service account to allow it to read and write pipeline artifacts from a cluster object storage bucket?
Answer: A. dedicated data connection role or equivalent S3 permissions
Pipelines authenticate to S3 storage using credentials defined within a data connection, which acts via the underlying role or equivalent permissions. Relying on a generic secret-reader role is insufficient because it lacks the specific S3 API authorization needed to read and write artifacts.
Q10. What configuration determines which cluster nodes can host OpenShift AI workloads based on specific labels or taints?
Answer: C. Node selectors and tolerations specified in the DSC / notebook configuration
Node selectors and tolerations specified in the DataScienceCluster and notebook configurations target specific labeled or tainted nodes. This is standard OpenShift scheduling logic, not a function of the Cluster Network Operator or OpenShift routing tables.
Q11. What is the role of Elyra inside the Red Hat OpenShift AI Jupyter environment?
Answer: C. Providing an interactive visual pipeline editor to build and run workflows
Elyra provides a visual pipeline editor within JupyterLab to help developers build and run data science workflows. It does not manage network load balancing or security scanning, focusing strictly on streamlining pipeline creation.
Q12. In OpenShift AI, what is a workbench?
Answer: D. A specialized pod running a development environment like JupyterLab
A workbench is a specialized pod running an interactive development environment like JupyterLab. It operates directly within the OpenShift cluster rather than functioning as a standalone virtual machine or a data visualization dashboard.
Q13. What parameter inside a Data Connection configuration defines the secret key used to access object storage?
Answer: A. AWS_SECRET_ACCESS_KEY
The AWS_SECRET_ACCESS_KEY parameter defines the secret key for accessing object storage in a Data Connection. OpenShift AI uses standard S3 naming conventions for its storage credentials, so expect these exact keys during exam tasks.
Q14. Which dashboard tab allows you to configure cluster-wide container image registries for OpenShift AI?
Answer: C. Settings > Notebook images
The Settings tab in the OpenShift AI dashboard provides the Notebook images section to manage custom container image registries. Cluster settings only handle core OpenShift platform configurations, not AI-specific notebook images.
Q15. What component state in the DataScienceCluster CR prevents the operator from managing or modifying the component configuration?
Answer: D. Unmanaged
Setting a component state to Unmanaged in the DataScienceCluster custom resource stops the operator from altering that specific component's configuration. Removed is used to uninstall a component, while Unmanaged freezes it.
Q16. What parameter defines the absolute idle duration before an OpenShift AI notebook cell or workbench is automatically culled?
Answer: B. cullingTimeout
The cullingTimeout parameter sets the absolute idle duration before an OpenShift AI workbench is automatically culled. This ensures idle resources are freed up, while other options do not exist in the dashboard settings.
Q17. Which OpenShift API group contains the DataScienceCluster Custom Resource Definition used to manage the RHOAI installation lifecycle?
Answer: A. opendatahub.io
The DataScienceCluster Custom Resource Definition belongs to the opendatahub.io API group, which manages the Red Hat OpenShift AI lifecycle. Remember that OpenShift AI is built directly on top of the upstream Open Data Hub project.
Q18. What field in the DataScienceCluster custom resource tracks the current operational state of the deployed applications?
Answer: A. status.phase
The status.phase field in the DataScienceCluster custom resource reflects the overall operational condition of the deployed components. ManagementState indicates intent, while status.phase shows the actual current runtime state.
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.