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 command validates that GPU labels are applied properly to nodes intended for OpenShift AI training workloads?. 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 command validates that GPU labels are applied properly to nodes intended for OpenShift AI training workl…
- WHICH WORKBENCH REPOSITORY CLONING APPROACH PREVENTS MERGE CONFLICTS ACCROSS TEAM WORKBENCHES CONTRIBUTING TO…
- WHICH STORAGE CONFIGURATION PROVIDES SHARED RUNTIME FILE SYSTEMS ENABLING MULTIPLE INDEPENDENT WORKBENCHES TO…
- What custom resource definition (CRD) is primarily used to provision a new Data Science Cluster?
- Which command enables retrieving the real-time status of the installed OpenShift AI controller status via CLI?
- How does a data science workbench authenticate to an external S3 storage bucket securely?
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. What command validates that GPU labels are applied properly to nodes intended for OpenShift AI training workloads?
Answer: C. oc get nodes –show-labels
Running the get nodes command with the appropriate flag displays node labels, allowing you to verify GPU tagging for scheduling. The other commands either query unrelated cluster resources or only display operator status.
Q2. WHICH WORKBENCH REPOSITORY CLONING APPROACH PREVENTS MERGE CONFLICTS ACCROSS TEAM WORKBENCHES CONTRIBUTING TO THE SAME GIT SOURCE?
Answer: B. Creating isolated feature branches for each data scientist work stream
Creating isolated feature branches prevents concurrent modifications from overwriting each other in shared workspaces. Sharing a single local workspace deployment directory introduces direct merge conflicts.
Q3. WHICH STORAGE CONFIGURATION PROVIDES SHARED RUNTIME FILE SYSTEMS ENABLING MULTIPLE INDEPENDENT WORKBENCHES TO COLLABORATE SIMULTANEOUSLY?
Answer: A. ReadWriteMany (RWX) Persistent Volume Claim
A ReadWriteMany persistent volume claim supports multiple independent pods mounting the volume simultaneously. ReadWriteOnce limits access to a single node, preventing concurrent multi-workbench collaboration.
Q4. What custom resource definition (CRD) is primarily used to provision a new Data Science Cluster?
Answer: C. DataScienceCluster
The DataScienceCluster custom resource provisions and configures the core OpenShift AI platform components. The DSCInitialization resource primarily handles foundational prerequisites before the main cluster deployment occurs.
Q5. Which command enables retrieving the real-time status of the installed OpenShift AI controller status via CLI?
Answer: C. oc get datasciencecluster -A
The DataScienceCluster custom resource reflects the actual installation state of OpenShift AI components. Querying the operators or pods directly won't give you the clean, top-level component health summary that the DSC provides.
Q6. How does a data science workbench authenticate to an external S3 storage bucket securely?
Answer: D. Using Data Connections backing an OpenShift Secret
Data Connections securely map endpoint and credential details directly into an OpenShift Secret. Hardcoding keys in notebooks or using plain text environment variables is explicitly tested against on the exam for poor security practices.
Q7. Which repository type cannot be linked directly via Data Connections as an automatic model registry sourcing pipeline?
Answer: B. Local USB drive connected to user laptop
Data Connections require network-accessible S3-compatible object storage endpoints. A local USB drive lacks a reachable URL and cannot be natively linked as a data source within the cluster.
Q8. What system step integrates a custom-built Docker container as an approved workbench category option in the UI?
Answer: A. Creating a Custom Notebook Image definition via the Cluster Settings Dashboard interface
Custom workbench images are exposed to users by creating a Custom Notebook Image definition in the dashboard. Modifying raw web assets or copying tar files manually is unsupported and breaks cluster updates.
Q9. Which file suffix identifies the compiled execution topology exported by Elyra to trigger an OpenShift Pipeline job?
Answer: A. .pipeline
Elyra exports visual pipeline canvases into compiled pipeline files using the dot pipeline extension. Standard YAML files are typically hand-written or generated by other CLI tools for basic task runs.
Q10. Which operational configuration prevents runaway resource depletion caused by idle notebook sessions within a cluster project namespace?
Answer: B. Notebook Idle Culler configurations settings
The Notebook Idle Culler is specifically designed to identify inactive notebook sessions and terminate them, freeing up cluster resources. ResourceQuotas merely restrict the total resource consumption allowed, but they do not actively cull idle pods.
Q11. Which specific configuration field within the DataScienceCluster CR represents an installed component's active state management?
Answer: A. managementState (Managed / Removed / Unmanaged)
The managementState field directly controls whether a component is actively managed by the operator. Expect to manipulate this field with oc edit or oc patch to enable or disable specific components on the exam.
Q12. Which S3 metadata specification identifies the repositories subfolder location sourcing a model directly in the Data Connection dialog?
Answer: C. Folder Path or Prefix parameter
The folder path or prefix parameter designates the exact subfolder within a bucket used to source the model. This field maps directly to object storage structures when configuring a data connection.
Q13. Which action occurs unconditionally to workbench data stored on permanent volumes upon rebooting the physical node sourcing workers?
Answer: A. Data remains persistent and intact on the external volume claim layer
Persistent storage guarantees data remains intact and available across node reboots via the underlying volume claim. Data loss only happens with ephemeral EmptyDir volumes, whereas workbenches use persistent volumes.
Q14. Where are the pipeline definitions and task layouts stored structurally when developing Elyra-based pipelines in OpenShift AI?
Answer: A. Within .pipeline files in the workspace
The Elyra extension saves visual pipeline layouts as .pipeline files directly within the Jupyter workspace. These files are later compiled into Kubernetes resources, rather than being stored in a database.
Q15. What container image registry pattern is required to add custom images to OpenShift AI in air-gapped deployments?
Answer: A. Mirrored OCI Registry via ICSP / IDMS
Disconnected environments require a locally mirrored Open Container Initiative registry with configured ImageContentSourcePolicy. Direct internet URLs or public registries will strictly fail in air-gapped cluster scenarios.
Q16. Which object category must be reconfigured if an administrator needs to restrict network egress for pipeline worker pods?
Answer: C. NetworkPolicy objects inside the target project namespace
NetworkPolicy resources restrict pod traffic entering or leaving a specific namespace. SecurityContextConstraints handle pod permissions and capabilities, but they do not manage network routing or egress filtering.
Q17. What kind of SSH credential pattern is recognized by the OpenShift AI user interface to allow private Git connections?
Answer: B. OpenSSH Private Key format stored as a cluster project secret
The dashboard explicitly requires an SSH private key formatted as a Kubernetes secret to sync private repositories. Plaintext passwords or PGP files are not supported natively by the connection UI.
Q18. WHAT ENVIRONMENT VARIABLE STORES THE OBJECT STORAGE ENDPOINT ACCESS KEY IN AN OPENSHIFT AI DATA CONNECTION CONFIGURATION?
Answer: D. AWS_ACCESS_KEY_ID
OpenShift AI data connections map directly to standard S3 environment variables, using AWS_ACCESS_KEY_ID for the access key. The other options are fabricated variable names and will not authenticate properly to the object store.
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.