HPE7-S02 Advanced HPE Compute Integrator Solutions Written Exam Questions and Answers

D is correct. Airflow pools provide a direct mechanism for controlling task concurrency. Each pool has a defined number of slots, and tasks assigned to that pool consume slots while they execute. Reducing the number of slots in the default pool therefore limits the number of tasks that can run simultaneously and places an upper bound on how aggressively Airflow can consume cluster resources across concurrent DAG activity.

Option A is not the supported control in this scenario; the Airflow UI does not serve as the general editor for executor-pod CPU and memory sizing. Option B is the opposite of the stated objective because enabling autoscaling can increase executor capacity in response to demand. Option C suggests changing framework values for executor resources, but HPE’s Airflow integration historically constrains those settings and the exam-tested method for limiting aggregate workload pressure is pool concurrency.

This is a scheduling-level safeguard rather than a Kubernetes quota. In a production environment, pool slots can be combined with platform quotas and workload resource requests to create layered control over resource consumption.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 4 “Airflow” and Module 7 resource troubleshooting; HPE AI Essentials Airflow documentation, pool slots and maximum simultaneous jobs.

D is correct for a customer-supplied, already-issued certificate. HPE AI Essentials presents its external TLS identity through the Istio ingress gateway, and the relevant certificate material is held in the ingress-cert secret in the istio-system namespace. When performing the manual replacement workflow, the administrator replaces the existing ingress certificate secret with one populated from the supplied certificate, private key, and CA chain according to HPE’s documented format.

Option A is incorrect because the certificate is not installed on a single UI virtual machine; HPE AI Essentials is a Kubernetes-based platform and ingress TLS is handled at the cluster ingress layer. Option B confuses certificate issuance with certificate installation. A ClusterIssuer is appropriate when cert-manager should request and renew a certificate, but the customer has already supplied the signed certificate and key. Option C incorrectly places TLS management in Keycloak.

Rev. 26.21 updated the certificate procedures for HPE AI Essentials 1.9 and later, so administrators should follow the version-specific HPE commands and secret format rather than an older generic Kubernetes recipe.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 3 “Replacing the Istio ingress gateway certificate”; HPE AI Essentials documentation, manual certificate replacement and ingress-cert access.

A is correct. VMware vSAN is designed to manage its storage devices directly so that it can create disk groups or storage pools, apply its own data-placement policies, and monitor device health and performance. On HPE ProLiant servers using a supported storage controller, vSAN data drives should therefore be presented in passthrough/HBA mode where the controller and vSAN design support it. This exposes the individual drives rather than hiding them behind a hardware RAID virtual disk.

RAID 1 and RAID 5 in options D and C would move redundancy and data placement into the controller, preventing vSAN from managing the devices as intended. Option B incorrectly treats erasure coding as a storage-controller mode. vSAN can apply software-defined RAID/erasure-coding policies at the datastore layer, but that is separate from configuring the physical controller.

The exact controller, firmware, driver, and device configuration must match the VMware Compatibility Guide and HPE vSAN design guidance. Passthrough is the architectural principle; supportability still depends on the validated hardware stack.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 8 “VMware vSAN on HPE solutions”; HPE and VMware vSAN configuration guidance.

D is correct. HPE Private Cloud AI can export observability data through an OpenTelemetry (OTel) endpoint. HPE AI Essentials retains Prometheus metrics for a defined local retention period, but organizations that require longer retention can register an OTel endpoint and send the telemetry to an external observability or storage platform with the retention policy they require. This decouples operational history from the local Prometheus retention window.

Option A does not describe the supported long-term metrics export mechanism. Option B would merely add another visualization component and would not solve durable metric retention; Grafana reads metrics but is not itself the long-term time-series store. Option C concerns alert output, which is a subset of operational events and cannot substitute for retained Prometheus time-series data.

This design is operationally important for capacity analysis, incident review, compliance evidence, and trend detection. The platform remains responsible for producing metrics, while the external OTel destination provides enterprise-scale storage and analytics.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Modules 3 and 7, monitoring and troubleshooting; HPE AI Essentials Software documentation, “Configuring Endpoints,” which describes OTel export for long-term metric storage.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 3 imported-framework communications and troubleshooting; HPE7-S02 Release Notes, June 15, 2026, framework-import updates and automatic Keycloak redirect URL handling.

B is correct. Livy provides a REST-based service for creating and managing interactive Apache Spark sessions. In HPE AI Essentials, notebook users who need interactive Spark execution rely on Livy so that notebook-side Spark magic or client commands can submit code to a managed Spark session without embedding a full Spark cluster process inside the notebook itself. Therefore the administrator should ensure Livy is installed and available through the platform’s Tools & Frameworks administration.

Option A is unnecessary because the platform can provide the supported notebook-to-Spark integration once Livy is present; simply building a custom image does not establish the backend session service. Option C assumes that an image name creates Spark connectivity, which it does not. Option D confuses the Spark Operator, used for Kubernetes-managed SparkApplication resources, with the interactive session mechanism requested in the question.

The exam distinction is between batch and interactive Spark. Spark Operator is well suited to declarative Spark applications; Livy enables interactive sessions from notebooks.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 4 “Using Spark” and Kubeflow notebooks; HPE AI Essentials Software documentation for Livy and interactive Spark access.

D is correct. NVIDIA NeMo Evaluator is designed to measure model and application quality against defined evaluation criteria, including safety-related dimensions. Current NeMo evaluation tooling includes safety evaluations for harmful-content generation, toxicity, bias, hate speech, and related robustness concerns. A customer who has fine-tuned a model and wants to determine whether it produces toxic output needs an evaluation service rather than a data-preparation or runtime-policy component.

NeMo Customizer is used to adapt or fine-tune models. NeMo Curator prepares training data through curation, filtering, and deduplication. NeMo Guardrails is used to apply runtime controls to conversational AI applications, such as restricting unsafe topics or validating inputs and outputs. Guardrails can help prevent unacceptable responses in production, but the question asks which microservice is used to evaluate whether the fine-tuned model is producing toxic content.

A sound AI lifecycle therefore separates customization, evaluation, and enforcement: customize the model, evaluate its behavior, and then apply appropriate guardrails and deployment controls.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 5 “NVIDIA NeMo deep dive”; NVIDIA NeMo Evaluator documentation, safety and security evaluation capabilities.

C is correct. The built-in KubeRay head service is normally exposed only inside the Kubernetes cluster, so an external client cannot simply call ray.init() against it by default. One supported way to provide interactive external access is to change kuberay-head-svc from ClusterIP to NodePort. After the service is exposed, the administrator identifies the assigned node port for the Ray client endpoint and connects to the cluster using a ray:// address.

Option A incorrectly states that external access is enabled by default. Port 10001 is the Ray client service port inside the service definition, but an external client still requires a reachable service exposure mechanism. Option D confuses the Ray dashboard port, commonly 8265, with the Ray client endpoint. Option B is too restrictive; a separate Ray cluster is not required merely to permit external interactive access.

Because NodePort expands network exposure, the change should be combined with appropriate firewalling, authentication, and access controls rather than treated as a convenience-only setting.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 4 “Using Ray”; HPE AI Essentials Software documentation, “Connecting to Ray Cluster.”

D is correct. Airflow is the HPE AI Essentials framework used to define, schedule, and execute directed acyclic graphs. A DAG expresses a workflow as tasks plus dependencies, so the Airflow interface is the appropriate place to demonstrate launching and monitoring a workflow. The UI exposes DAG state, task instances, execution history, schedules, and operational details that make the orchestration behavior visible during a customer demonstration.

NVIDIA AI Enterprise is the broader enterprise AI software suite and includes components such as NIM and NeMo, but it is not the HPE AI Essentials DAG scheduler. MLflow is used for experiment tracking and model lifecycle activities. Ray is a distributed computing framework used for scalable Python and AI workloads; it does not provide the Airflow DAG-management interface requested here.

A useful demonstration can show both control and observability: trigger a DAG, inspect task dependency order, observe task status, and review logs or retries. That makes the distinction between workflow orchestration and the underlying execution frameworks clear.

References/topics: Advanced HPE Compute Solutions, Rev. 26.21, Module 4 “Airflow” and the objective “Demonstrate HPE AI Essentials”; HPE AI Essentials Software Airflow documentation.