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: To expose a deployed machine learning model to applications outside the OpenShift cluster, what resource does OpenShift . 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
- To expose a deployed machine learning model to applications outside the OpenShift cluster, what resource does…
- Which technology provides the foundation for model serving endpoints scalability in KServe?
- Which Custom Resource contains the configuration parameters for integrating a custom inference engine into th…
- What type of secret is generated to authenticate OpenShift AI pipelines with an external Git provider?
- What happens to data stored in a workbench's local directory if the workbench pod is restarted?
- Which service mesh framework is a pre-requisite dependency for running the KServe single-model serving platfo…
Answers and explanations
Tap a question to expand the answer and the exam reasoning. Try to commit to your own pick first.
Q1. To expose a deployed machine learning model to applications outside the OpenShift cluster, what resource does OpenShift AI generate?
Answer: A. An Ingress or OpenShift Route
Serving runtimes in OpenShift AI expose machine learning models to external applications using standard Kubernetes networking resources like Ingress or OpenShift Routes. ClusterIP services only provide internal cluster access, making them unsuitable for direct external consumption.
Q2. Which technology provides the foundation for model serving endpoints scalability in KServe?
Answer: B. Knative Serving framework
KServe relies on the Knative Serving framework to provide advanced scaling and traffic management for model serving endpoints. HAProxy configurations alone cannot handle the scale-to-zero capabilities required by KServe.
Q3. Which Custom Resource contains the configuration parameters for integrating a custom inference engine into the OpenShift AI serving layer?
Answer: B. ServingRuntime
The ServingRuntime custom resource defines the configuration parameters needed to integrate a custom inference engine into the OpenShift AI serving layer. AcceleratorProfile handles hardware scheduling, while DataScienceCluster manages operator components.
Q4. What type of secret is generated to authenticate OpenShift AI pipelines with an external Git provider?
Answer: B. Basic authentication or SSH key secret
OpenShift AI pipelines use basic authentication or SSH key secrets to securely authenticate with external Git providers for code retrieval. Service account tokens are meant for internal API access, whereas Git integration demands explicit credential management.
Q5. What happens to data stored in a workbench's local directory if the workbench pod is restarted?
Answer: C. Data is preserved if backed by a PersistentVolumeClaim
Data in a workbench survives pod restarts only when mapped to a persistent volume claim, which decouples storage from the pod lifecycle. Without a claim, the ephemeral container storage is destroyed when the pod terminates.
Q6. Which service mesh framework is a pre-requisite dependency for running the KServe single-model serving platform in RHOAI?
Answer: B. OpenShift Service Mesh (Istio)
KServe requires OpenShift Service Mesh, based on Istio, to handle advanced networking and routing for single-model serving. Linkerd and Consul are not compatible with the strict RHOAI operator dependencies for KServe.
Q7. What status property verifies that all required sub-components of Red Hat OpenShift AI are operational after an upgrade?
Answer: B. The phase field showing Ready/Reconciled in the DataScienceCluster status resource output
The DataScienceCluster custom resource indicates operational health when its phase field shows Ready or Reconciled. This confirms all sub-components finished reconciling successfully, making it the primary validation step.
Q8. What parameter in a Data Connection YAML configuration determines the name of the S3 bucket containing the target machine learning assets?
Answer: A. AWS_S3_BUCKET
The AWS_S3_BUCKET parameter within a Data Connection YAML specifies the exact S3 bucket holding the target machine learning assets. The other parameters do not directly map to the bucket name needed for S3 connectivity.
Q9. What Kubernetes resource type is utilized to define the data structures for managing custom model serving runtimes?
Answer: A. ConfigMap or ServingRuntime Custom Resource Definitions
Custom serving runtimes in OpenShift AI are defined using the ServingRuntime custom resource, which dictates the pod templates and configurations. ConfigMaps are commonly used alongside these definitions to store necessary runtime scripts or environment configurations.
Q10. What type of data traffic passes through the KServe data plane interface?
Answer: A. Inference request payloads and corresponding model prediction responses
The KServe data plane handles the actual inference payloads, routing client prediction requests to the deployed models and returning the outputs. Administrative configurations or source code transfers happen outside the inference data plane interfaces.
Q11. To scale down model serving instances to zero when no traffic is detected, which dependency must be configured and running properly?
Answer: A. OpenShift Serverless (Knative)
Scaling models to zero requires OpenShift Serverless, which provides the Knative Serving infrastructure to manage scale-to-zero functionality. Standard Service Mesh alone cannot scale pods to zero without the Knative abstraction managing the network endpoints.
Q12. Which interface inside an OpenShift AI workbench allows creating pipeline files visually?
Answer: A. Elyra Pipeline Editor
The Elyra Pipeline Editor provides the visual drag-and-drop interface used to create pipeline files within the workbench. While JupyterLab is the underlying environment, the Extension Manager merely handles installing tools rather than authoring pipelines.
Q13. Where are the pipeline artifacts, such as intermediate datasets or model weights, typically stored during an Elyra pipeline run?
Answer: B. In an object storage bucket defined via Data Connection
Pipeline artifacts like datasets and weights are typically stored in an object storage bucket defined by a Data Connection. ConfigMaps and Git repositories are meant for configuration and code, not large binary model artifacts.
Q14. Which OpenShift AI capability allows multiple distinct machine learning models to share the exact same runtime container pod?
Answer: C. ModelMesh Multi-Model Serving Runtime
ModelMesh is the multi-model serving runtime designed specifically to share a single container pod among multiple distinct machine learning models. KServe, by contrast, deploys each model in its own dedicated single-model pod.
Q15. What happens to data stored inside a workbench's Persistent Volume Claim (PVC) when the workbench is stopped or culled?
Answer: D. The data is preserved and reattached when restarted
Workbench storage is backed by a Persistent Volume Claim, meaning data is preserved and automatically reattached whenever the workbench restarts. This prevents data loss during standard culling operations or intentional scale-downs.
Q16. Which tool inside the OpenShift AI workbench environment enables visual execution and building of data science pipelines?
Answer: B. Elyra Pipeline Editor
The Elyra Pipeline Editor is the visual tool within the OpenShift AI workbench used to build and execute data science pipelines. While VS Code is available as an IDE, Elyra provides the specific drag-and-drop pipeline interface.
Q17. Which operator is required to install Red Hat OpenShift AI on an OpenShift cluster?
Answer: C. Red Hat OpenShift AI Operator
The Red Hat OpenShift AI Operator is the required operator to install and manage OpenShift AI on a cluster. Other operators like GitOps might support operational workflows but are not required for the core installation.
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.