Advanced HPE Compute Integrator Solutions v1.0

Page:    1 / 3   
Exam contains 40 questions
Expand All

You are proposing HPE Private Cloud AI to a financial institution and want to demo the solution running a complex AI-enabled security application.
What can help you to accelerate the development of the demo?

  • A. Finding sample Kubeflow pipelines in the HPE AI Essentials shared volume
  • B. Looking for an appropriate NVIDIA blueprint in the NGC catalog
  • C. Using tools within NVIDIA AI Enterprise Base Command Manager
  • D. Searching for tutorials in the HPE AI Essentials documentation


Answer : B

You need to connect several existing HPE Cray XD servers to a new HPE Slingshot network featuring HPE Slingshot ToR switches. The servers have existing 200 Gbps Ethernet NICs.
What is required?

  • A. Replacing the NICs with InfiniBand NDR NICs
  • B. Checking whether the NICs are qualified for connecting to the HPE Slingshot switch
  • C. Replacing the NICs with HPE Slingshot NICs
  • D. Installing an LNet router between the servers and HPE Slingshot switches


Answer : B

What is one difference between a Class 1 and a Class 3 HPE Slingshot topology?

  • A. The Class 1 topology supports more compute node NICs per switch that the Class 3 topology.
  • B. The Class 1 topology scales to a larger number of compute node NICs than the Class 3 topology.
  • C. The Class 1 topology supports a higher ratio of L1 and L2 links versus L0 links than the Class 3 topology.
  • D. The Class 1 topology supports additional adaptive routing technologies beyond those supported with the Class 3 topology.


Answer : A

A customer needs to deploy a NIM LLM with HPE Machine Learning Inference Software running on HPE Private Cloud AI (AI Essentials 1.6.1).
What is one option for the image for serving the model?

  • A. A custom image stored in the customer’s choice of registry
  • B. An image available in the public NGC registry
  • C. An image stored in the HPE Private Cloud AI registry
  • D. A BentoML image provided by HPE Machine Learning Inference Software


Answer : C

A customer is running SUSE Rancher on HPE ProLiant DL servers and HPE Alletra arrays.
How should you install the HPE CSI Driver for Kubernetes?

  • A. Download the latest Helm chart provided by HPE for the driver. Install that chart.
  • B. Obtain the latest Ansible playbook for the driver from the HPE GitHub site. Run the playbook.
  • C. Use the SUSE Rancher Operators catalog to find the HPE driver. Install the driver from there.
  • D. Install an HPE Data Ops Manager plug-in for SUSE Rancher. Use the plug-in to deploy the CSI driver.


Answer : C

How do workloads running within HPE AI Essentials authenticate each other?

  • A. Using certificates issued by the Istio ingress gateway built into HPE AI Essentials
  • B. Using SVIDs that an Istio proxy helps them obtain from a built-in SPIRE server
  • C. Using Kubernetes secrets auto-generated by Kyverno mutating admission controllers when workloads spin up
  • D. Using self-signed certificates generated when HPE installs the solution


Answer : B

A customer has HPE ProLiant DL Gen12 servers managed by HPE Compute Ops Management and acting as VMware vSphere ESXi hosts.
How can you handle updating the servers’ firmware and OS together?

  • A. Deploy HPE OneView for vCenter Orchestrator. Use a workflow to automate the firmware and OS updates.
  • B. Deploy HPE OneView for vCenter (0V4VC) and integrate it with vCenter and HPE Compute Ops Management. Then manage the updates with 0V4VC cluster profiles.
  • C. Set up an integration between HPE Compute Ops Management and VMware Cloud Foundation Operations. Use the Operations workflows to automate the updates.
  • D. Set up an integration between HPE Compute Ops Management and VMware vCenter. Then manage the updates from vSphere Lifecycle Manager.


Answer : D

An HPE Private Cloud AI customer wants to replace the self-signed certificate that users see when they access HPE AI Essentials. The customer wants to reduce maintenance efforts by having the solution auto-renew the certificate.
What is part of the implementation process?

  • A. Creating a ClusterIssuer object in HPE AI Essential’s cert-manager namespace
  • B. Creating an authorization policy in HPE AI Essential’s cert-manager namespace
  • C. Creating a secret in HPE AI Essential’s istio-system namespace that includes a certificate, private key, and CA certificate
  • D. Creating an ingress gateway in HPE AI Essential’s istio-system namespace


Answer : A

Which benefit does HPE Rear Door Heat Exchanger (RDHX) provide?

  • A. Decreasing energy costs through the use of liquid-to-air cooling
  • B. Allowing server chassis to tolerate higher temperatures than those for which they are rated
  • C. Providing the efficiencies of liquid cooling to facilities without water available at the rack
  • D. Reducing costs by removing the need for heat sinks and fans within servers


Answer : A

An AI team is using Spark Applications to analyze and format data for AI fine-tuning. The team would like to automate running the applications in a series.
Which component of HPE AI Essentials can help?

  • A. Katib
  • B. Airflow
  • C. Ray Tune
  • D. MLflow


Answer : B

An HPE Private Cloud AI User wants to use KServe to run a NIM that is available in NGC.
What is one prerequisite?

  • A. Creating an NGC registry in HPE Machine Learning Inference Software that includes credentials for a NVIDIA AI Enterprise account
  • B. Importing NVIDIA AI Enterprise into HPE AI Essentials using a Helm chart
  • C. Adding KServe decorators to the code that launches the NIM
  • D. Creating an NGC API Key and specifying it in a secret in the user’s namespace


Answer : D

An HPE Private Cloud AI customer has a model deployed with HPE Machine Learning Inference Software (MLIS). You have helped the customer further fine-tune that model. Now the customer would like to test the new model by directing 10% of traffic to it.
What should you do?

  • A. Add a new version of the model to the existing deployment as a canary roll out, setting the desired percentages per version.
  • B. Create a second deployment with the new model in MLIS. Use HPE Private Cloud AI’s Canal service to set up routing rules to each deployment.
  • C. Migrate the deployment to KServe and use autoscaling to set the 10% value for the new model.
  • D. Edit the existing deployment and use the advanced options to set up routing rules to the two model versions.


Answer : A

Which correctly describes an aspect of the AI software stack for HPE Cray XD systems?

  • A. The systems use HPE GreenLake cloud for multi-systems management.
  • B. The systems use HPE Performance Manager Cluster for the AI development environment.
  • C. The systems use a combined firmware and OS, developed by HPE.
  • D. The systems use RedFish-based firmware and a Linux OS.


Answer : D

You are deploying several HPE ProLiant DL Gen12 servers to act as ESXi hosts.
How can you help to ensure the servers pass TPM attestation?

  • A. Enable UEFI Secure Boot.
  • B. Enable encryption on all self-encryting drives (SEDs).
  • C. Order TAA-compliant models.
  • D. Enable the iLO Service Port.


Answer : A

A customer requires accessing the HPE AI Essentials KubeRay cluster interactively (with ray-init) from an external client.
What should you explain?

  • A. External access is enabled by default at the Ray dashboard UI. The external client should direct the ray-int requests to TCP port 10001.
  • B. External access is not supported with the built-in KubeRay cluster. The customer should deploy a new cluster to achieve this access.
  • C. External access is not available by default, but can be achieved by, for example, changing the kuberay-head-svc to a Node Port service.
  • D. External access is enabled by default at the Ray dashboard UI. The external client should direct the ray-int requests to TCP port 8265.


Answer : C

Page:    1 / 3   
Exam contains 40 questions
Expand All

Talk to us!

Have any questions or issues ? Please dont hesitate to contact us

[email protected]

Certlibrary.com is owned by MBS Tech Limited: Room 1905 Nam Wo Hong Building, 148 Wing Lok Street, Sheung Wan, Hong Kong. Company registration number: 2310926
Certlibrary doesn't offer Real Microsoft Exam Questions. Certlibrary Materials do not contain actual questions and answers from Cisco's Certification Exams.
CFA Institute does not endorse, promote or warrant the accuracy or quality of Certlibrary. CFA® and Chartered Financial Analyst® are registered trademarks owned by CFA Institute.
Terms & Conditions | Privacy Policy | Amazon Exams | Cisco Exams | CompTIA Exams | Databricks Exams | Fortinet Exams | Google Exams | Microsoft Exams | VMware Exams