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 is the primary benefit of using an ImageStream to manage notebook images in OpenShift AI?. 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 is the primary benefit of using an ImageStream to manage notebook images in OpenShift AI?
- Which OpenShift AI component manages the lifecycle of distributed AI workloads using frameworks like Ray or C…
- Which interface is used by data scientists as the central hub to create projects, launch workbenches, and vie…
- What parameter in a Workbench configuration determines the persistent disk space allocated for user code and…
- Which environment variable is automatically populated when a Data Connection is attached to an OpenShift AI W…
- Which protocol is typically used to send inference requests to a model deployed on ModelMesh?
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. What is the primary benefit of using an ImageStream to manage notebook images in OpenShift AI?
Answer: B. It allows automatic updates and tracking of container images within the cluster
Using an ImageStream enables automatic updates and tracking of container tags within the cluster, streamlining notebook image management. It does not compile Python code; it simply points to the underlying container registry layer.
Q2. Which OpenShift AI component manages the lifecycle of distributed AI workloads using frameworks like Ray or CodeFlare?
Answer: C. Distributed Workloads / Training operator
Distributed AI workloads utilizing Ray or CodeFlare are managed by the distributed workload components and the Training operator. Single model serving handles inference endpoints rather than orchestrating distributed training jobs.
Q3. Which interface is used by data scientists as the central hub to create projects, launch workbenches, and view connections?
Answer: C. Red Hat OpenShift AI Dashboard
The Red Hat OpenShift AI Dashboard serves as the primary hub for data scientists to manage projects, launch workbenches, and configure connections. The OpenShift Container Platform Web Console handles cluster administration rather than user workflows.
Q4. What parameter in a Workbench configuration determines the persistent disk space allocated for user code and notebook dependencies?
Answer: A. Volume size
The volume size parameter dictates the capacity of the persistent volume claim attached to the workbench. This ensures user code, datasets, and notebook dependencies persist across pod restarts, unlike replica counts which manage scaling.
Q5. Which environment variable is automatically populated when a Data Connection is attached to an OpenShift AI Workbench?
Answer: B. AWS_ACCESS_KEY_ID
Attaching a Data Connection to a workbench automatically injects storage credentials as environment variables, including AWS_ACCESS_KEY_ID. This mechanism simplifies accessing S3-compatible storage without hardcoding secrets into your notebooks.
Q6. Which protocol is typically used to send inference requests to a model deployed on ModelMesh?
Answer: A. gRPC or HTTP/REST
Models deployed with ModelMesh are typically queried using standard gRPC or HTTP REST protocols. Legacy protocols like FTP or SOAP are not supported for modern inference workloads in OpenShift AI.
Q7. What happens to data stored outside the /opt/app-root/src directory in a standard OpenShift AI notebook when the pod restarts?
Answer: B. It is lost if not backed up or mounted to a Persistent Volume
Data outside the mounted persistent volume is ephemeral, so it disappears when the pod restarts. Standard workbench images mount persistent storage to /opt/app-root/src, making any files saved outside this directory strictly temporary and lost during container recreation.
Q8. Which environment variable does the RHOAI notebook controller inject into a workbench pod to identify the logged-in user account?
Answer: A. JUPYTERHUB_USER
The notebook controller injects JUPYTERHUB_USER to identify the logged-in user account. The workbench leverages the underlying JupyterHub spawner mechanism, making this the standard environment variable rather than a custom Red Hat OpenShift AI specific variable.
Q9. To allow external applications to hit a serving runtime endpoint securely, which OpenShift network component must be generated?
Answer: C. Route
An OpenShift Route exposes an internal service to external traffic, making the model endpoint securely accessible. While a ClusterIP provides internal discovery, the Route handles external hostname routing and optional TLS termination.
Q10. How does an administrator add a new runtime engine choice into the Model Serving dropdown selection list globally?
Answer: D. By creating a cluster-scoped ServingRuntime or ClusterServingRuntime custom resource
Creating a cluster-scoped ServingRuntime custom resource globally registers a new runtime engine in the serving dashboard. Namespace-scoped resources limit choices to specific projects, whereas the cluster scope ensures the runtime appears as a global dropdown option.
Q11. Where are the access credentials for a newly configured Data Connection securely stored inside an OpenShift Data Science Project?
Answer: A. Secret
Kubernetes Secrets securely store sensitive access credentials like S3 keys. When a Data Connection is created, the operator generates a Secret rather than a ConfigMap, allowing workbenches and runtimes to mount these credentials as environment variables safely.
Q12. Which Custom Resource must be created in the project namespace to instruct KServe to deploy a specific model file?
Answer: D. InferenceService
KServe uses the InferenceService custom resource to deploy and manage a specific model file. This resource defines the model configuration and references the chosen serving runtime, automatically orchestrating the necessary pods and network services.
Q13. Which storage protocol is officially recommended for sharing pipeline artifacts within OpenShift AI Pipelines?
Answer: C. S3-compatible Object Storage
OpenShift AI Pipelines rely on S3-compatible object storage to share artifacts between pipeline steps. Block storage options fail because pipeline pods are ephemeral and require centralized access.
Q14. Which OpenShift AI component is responsible for orchestrating repetitive and automated machine learning workflows?
Answer: C. OpenShift AI Pipelines
OpenShift AI Pipelines orchestrates automated machine learning workflows. JupyterLab only provides interactive workspaces, while ModelMesh focuses exclusively on runtime model serving rather than workflow automation.
Q15. What technology allows vLLM to significantly reduce memory fragmentation and maximize throughput when serving large language models?
Answer: B. PagedAttention algorithms
vLLM uses PagedAttention to reduce memory fragmentation and maximize throughput for large language models. This algorithm manages attention keys and values efficiently, unlike standard memory paging mechanisms.
Q16. Which framework is optimized for high-density, multi-model serving within Red Hat OpenShift AI?
Answer: B. ModelMesh Serving
ModelMesh is explicitly designed for high-density, multi-model serving in OpenShift AI. KServe Raw operates as a single-model runtime, lacking the intelligent scaling and memory management of ModelMesh.
Q17. Which open-source project serves as the upstream codebase for Red Hat OpenShift AI?
Answer: D. Kubeflow / Open Data Hub
Open Data Hub serves as the upstream project for Red Hat OpenShift AI. Hadoop and TensorFlow are analytics tools, while OpenStack focuses entirely on infrastructure provisioning.
Q18. How do you update the Red Hat OpenShift AI operator to the latest minor version?
Answer: B. Change the subscription channel or update approval strategy in the OperatorHub console
Operator updates are managed through the OperatorHub subscription channel and approval strategy. The cluster upgrade command only updates the OpenShift platform itself, not individual operators like 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.