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 configuration ConfigMap stores the list of allowed ad-hoc notebook sizes displayed in the dashboard drawer?. 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 configuration ConfigMap stores the list of allowed ad-hoc notebook sizes displayed in the dashboard dra…
- WHAT TEMPLATE OBJECT REGISTERS A CUSTOM JUPYTER NOTEBOOK IMAGE GLOBALLY IN OPENSHIFT AI?
- WHICH MECHANISM ALERTS SRE TEAMS WHEN AN OPENSHIFT AI ENDPOINT MODEL STARTS REJECTING MORE THAN 10 PERCENT OF…
- WHAT TYPE OF PERSISTENT STORAGE SCHEME ALLOWS DELETING A WORKBENCH SPEC WITHOUT LOSING LOCAL PERSISTENT WORKS…
- Which labeling criteria specifies that a namespace is recognized as an active OpenShift AI data science proje…
- What environment variable in a Jupyter pod captures the logged-in OpenShift user account name?
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. Which configuration ConfigMap stores the list of allowed ad-hoc notebook sizes displayed in the dashboard drawer?
Answer: D. odh-dashboard-config
Custom notebook sizes are defined within the odh-dashboard-config ConfigMap. While generic resource names might seem correct, the dashboard specifically reads environment settings from this core configuration object.
Q2. WHAT TEMPLATE OBJECT REGISTERS A CUSTOM JUPYTER NOTEBOOK IMAGE GLOBALLY IN OPENSHIFT AI?
Answer: C. ImageStream
An ImageStream is the standard OpenShift template object used to register custom Jupyter notebook images globally. Providing a custom notebook image requires creating an ImageStream and pointing your workbench to it.
Q3. WHICH MECHANISM ALERTS SRE TEAMS WHEN AN OPENSHIFT AI ENDPOINT MODEL STARTS REJECTING MORE THAN 10 PERCENT OF TOTAL REQUESTS?
Answer: B. Prometheus Alertmanager rules matching ServiceMonitor metrics
Prometheus Alertmanager rules matching ServiceMonitor metrics correctly alert teams of high rejection rates. While you can manually inspect route statuses or pod logs, configuring Alertmanager provides automated, real-time monitoring for endpoint failures.
Q4. WHAT TYPE OF PERSISTENT STORAGE SCHEME ALLOWS DELETING A WORKBENCH SPEC WITHOUT LOSING LOCAL PERSISTENT WORKSPACE JUPYTER PROJECTS?
Answer: A. Detached Persistent Volume Claims (PVC) configurations
Using detached Persistent Volume Claims ensures data persists even if the workbench specification is deleted. This configuration keeps the PVC intact so users can reattach it later to recover their local Jupyter projects.
Q5. Which labeling criteria specifies that a namespace is recognized as an active OpenShift AI data science project?
Answer: C. opendatahub.io/dashboard: "true"
The OpenShift AI dashboard identifies managed namespaces by the opendatahub.io/dashboard=true label. For the exam, remember that UI-provisioned data science projects automatically apply this specific label to integrate with the dashboard.
Q6. What environment variable in a Jupyter pod captures the logged-in OpenShift user account name?
Answer: B. JUPYTER_USER
The JUPYTER_USER environment variable injects the authenticated OpenShift username into the running workbench pod. This variable is heavily relied upon during custom notebook configurations and when scripting shared volume access dynamically.
Q7. When saving ML models to be compatible with the OpenVINO runtime, which format preserves highest load speed performance?
Answer: D. OpenVINO IR (Intermediate Representation) XML/BIN
The OpenVINO Intermediate Representation, saved as paired XML and BIN files, is the native format optimized for fastest inference loading. While ONNX is supported, converting it to IR first guarantees optimal runtime compatibility.
Q8. Where does OpenShift AI store metric data and user configuration graphs by default?
Answer: B. Prometheus instance in redhat-ods-monitoring namespace
OpenShift AI utilizes a dedicated Prometheus instance running in the redhat-ods-monitoring namespace for internal metrics. Standard cluster monitoring is kept separate, so remember the specific namespace for dashboard components.
Q9. What method updates internal datasets cached inside a running workbench volume without disrupting the session environment live state?
Answer: B. Utilizing a Data Connection s3 sync routine from an external object store
Using an AWS CLI S3 sync command from within the workbench terminal updates datasets seamlessly using a Data Connection. Direct oc rsync operations are discouraged compared to native S3 integrations configured via environment variables.
Q10. WHICH CUSTOM RESOURCE (CR) INITIATES THE ACTUAL PROVISIONING OF THE OPENSHIFT AI CONTROL PLANE?
Answer: B. DataScienceCluster
The DataScienceCluster custom resource initiates the actual provisioning of the OpenShift AI control plane components. Installing the operator alone does nothing until you create this specific resource to trigger the component deployments.
Q11. When connecting a Jupyter Notebook to an object storage bucket, which credential type is typically injected via an OpenShift Secret?
Answer: A. AWS Access Key ID and Secret Access Key
S3-compatible object storage credentials, specifically the Access Key ID and Secret Access Key, are typically stored in an OpenShift Secret. The workbench then mounts this Secret to authenticate your notebook environment.
Q12. Which service exposes an internally deployed KServe predictor to downstream microservices acting in the same namespace?
Answer: D. ClusterIP Service
A ClusterIP service is correct because it provides internal networking exposure restricted to resources within the same namespace. Options like LoadBalancer or NodePort expose workloads externally, which violates the requirement for strictly internal downstream access.
Q13. WHICH LOGIC ENABLES ZERO-DOWNTIME ROLLOUTS WHEN RE-DEPLOYING UPDATED TRAINED MODELS ON KSERVE PLATFORMS?
Answer: B. Blue-Green or Canary Routing mechanisms via Knative Traffic Splitting routing configurations
Blue-Green or Canary routing via Knative traffic splitting is correct because KServe leverages Knative to manage gradual rollouts and zero-downtime updates. Standard Kubernetes rolling updates lack the ingress traffic manipulation required for advanced model deployments.
Q14. WHICH COMMAND-LINE TOOL CAN BE GENERALLY USED TO INTERACT WITH RED HAT OPENSHIFT AI RESOURCES PROGRAMMATICALLY?
Answer: A. oc CLI
The OpenShift CLI is the primary tool for interacting programmatically with Red Hat OpenShift AI resources. While specialized tools exist for specific tasks, the standard CLI handles general platform operations and custom resource management.
Q15. Which specific cluster component accelerates generative AI and deep learning model training in an OpenShift AI environment?
Answer: B. NVIDIA GPU Operator
The NVIDIA GPU Operator provisions and manages the hardware drivers required to accelerate deep learning and generative AI workloads. Operators handling networking or storage cannot provide the necessary compute acceleration for training models.
Q16. Which object enforces compute resource quotas across notebook pods within a data science project?
Answer: A. ResourceQuota
A ResourceQuota constrains the total aggregate compute resources consumed within a specific namespace. LimitRange is the strongest distractor, but it enforces per-pod minimum and maximum bounds instead of overall project totals.
Q17. What is the prerequisite operator required before configuring the OpenShift AI monitoring stack?
Answer: A. Red Hat OpenShift Cluster Monitoring Operator
The platform monitoring operator must be installed first because OpenShift AI metrics rely on the underlying Prometheus stack. Service mesh and pipelines are unrelated distractors that do not provide the required metrics ingestion backend.
Q18. Which type of encryption protects AI training data connection traffic traveling from workbenches to S3 buckets?
Answer: A. HTTPS TLS Encryption
HTTPS TLS encryption secures data in transit between workbenches and S3 compatible object storage. Disk encryption only protects data at rest, and unencrypted channels fail to provide the required wire-level security for sensitive payloads.
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.