Test
ML Platform Engineer
Este test tiene tiempo límite, así que se abre en el reproductor de siempre.
Preguntas del test (39)
1How can sensitive information like Azure API keys or passwords be securely managed in Terraform?
- AStore them directly in Terraform configuration files for easy access.
- BUse Azure Key Vault to store sensitive values and retrieve them during execution.
- CMark variables as sensitive in the Terraform configuration.
- DUse environment variables to pass sensitive information securely.
- ECommit the sensitive information to version control for team collaboration.
2What are best practices for using Terraform with an Azure remote backend?
- AStore the Terraform state in an Azure Blob Storage with encryption enabled.
- BUse Azure Resource Manager (ARM) policies to restrict access to the storage account.
- CEnable state locking by configuring Azure Storage to use Blob lease locks.
- DStore the Terraform state file on the local machine to reduce latency.
- EDisable versioning for the state file to simplify updates.
3Why is Terraform state management critical in an Azure infrastructure?
- AIt keeps track of the resources Terraform manages and their current state in Azure.
- BIt ensures idempotency by comparing the desired state with the actual state in Azure.
- CIt removes the need for version control by storing resource definitions directly in the state file.
- DIt prevents resource duplication by identifying existing Azure resources.
4How does Terraform manage dependencies between Azure resources?
- ABy manually specifying the order of resource creation in the configuration.
- BBy analyzing implicit references between resources in the configuration.
- CBy using resource outputs to reference attributes of other resources.
- DBy constructing a dependency graph based on the configuration.
- EBy always creating resources in alphabetical order.
5What are some common best practices for organizing Terraform configurations in an Azure environment?
- AUse separate workspaces or directories for development, staging, and production environments.
- BOrganize reusable components into Terraform modules.
- CWrite all configurations in a single file to simplify management.
- DStore secrets in Azure Key Vault instead of hardcoding them.
- EAvoid using remote backends for state files to prevent complexity.
6Which Terraform features help ensure security and compliance in Azure?
- AEnabling encryption for remote state stored in Azure Blob Storage.
- BUsing Azure Policy in conjunction with Terraform to enforce compliance.
- CMarking sensitive variables with sensitive = true.
- DCommitting sensitive state files to private repositories.
- EUsing public GitHub repositories to manage Terraform configurations.
7How can you automate Terraform workflows in Azure DevOps?
- AUse Azure DevOps Pipelines to trigger Terraform plan and apply commands.
- BStore Terraform state in an Azure Blob Storage backend with access restricted to the pipeline.
- CUse service principals to authenticate Terraform with Azure resources.
- DRun Terraform commands manually to ensure human oversight.
- EUse local state files to avoid pipeline complexity.
8How does Terraform handle parallelism during the creation of Azure resources?
- AIt creates resources in parallel where no explicit or implicit dependencies exist.
- BIt uses the dependency graph to determine which resources can be created concurrently.
- CIt waits for all resources to be created one at a time for better error handling.
- DIt ignores dependencies and creates resources randomly for faster execution.
- EIt allows for performance optimization by adjusting the parallelism flag.
9Which of the following practices help avoid "orphaned" resources in Terraform for Azure?
- ARegularly running terraform destroy for resources removed from the configuration.
- BUsing Terraform's state file to identify unused resources in Azure.
- CImplementing lifecycle rules to control the behavior of resource deletions.
- DAvoiding updates to Terraform configurations to prevent mismatches.
- EIgnoring state file updates when applying changes to the infrastructure.
10What are some benefits of using modules in Terraform for Azure?
- APromotes reusability of infrastructure configurations across multiple projects.
- BSimplifies complex configurations by encapsulating related resources.
- CEnables consistency across environments by using shared definitions.
- DAutomatically prevents errors without requiring additional testing.
- EEliminates the need for variable declarations.
11A developer reports that a new Azure virtual machine (VM) deployment failed during a Terraform apply operation. The error message states that the subnet is full.
- AAnalyze the Error:
Review the Terraform apply logs to confirm the specific error message.
Check if the error is caused by insufficient IP addresses in the subnet. - BVerify Subnet Configuration:
Go to the Azure Portal and navigate to the Virtual Network where the subnet resides.
Check the subnet’s IP range (CIDR) and the number of IP addresses already allocated.
Confirm if there are any unused resources (e.g., VMs, NICs, load balancers) consuming IP addresses. - CIdentify Active Resources:
Use az network nic list or the Azure Portal to list all network interfaces associated with the subnet.
Identify unused resources and delete them if they are no longer required. - DExpand the Subnet:
If the subnet is indeed full, expand its CIDR range (if allowed by the network design).
Update the Virtual Network configuration in Azure. - EUpdate Terraform Configuration:
If expanding the subnet is not feasible, create a new subnet with a larger range in the Terraform configuration.
Refactor Terraform scripts to associate the VM with the new subnet. - FApply Terraform Changes:
Re-run terraform plan to confirm the configuration changes.
Execute terraform apply to redeploy the resources. - GPrevent Future Issues:
Plan subnet CIDR ranges more strategically to accommodate growth.
Regularly monitor IP address usage and clean up unused resources. - HAnalyze the Error:
Review the Terraform logs for unrelated warning messages.
Assume the issue is due to a Terraform bug without further checks. - IVerify Subnet Configuration:
Check the subnet configuration by looking at its name only.
Confirm that unused resources are consuming IP addresses as expected. - JIdentify Active Resources:
Use az network nic list to delete all network interfaces in the subnet without verification.
Rely on manual checks without validating unused resources.
- AAnalyze the Error:
12Your team uses Terraform for Azure infrastructure but has reported a security incident where sensitive information was leaked.
What steps would you take to ensure sensitive information is not exposed in the future?- AImmediate Actions:
Revoke access to compromised credentials or secrets.
Rotate the affected keys, tokens, or credentials immediately. - BConduct a Post-Mortem:
Identify how the sensitive information was leaked (e.g., plain-text secrets in code, state files, or logs).
Analyze Terraform state files (terraform.tfstate) for exposed secrets. - CSecure Sensitive Data:
Use Terraform's Vault integration or Azure Key Vault to manage sensitive information securely.
Encrypt Terraform state files using remote state storage with secure backends (e.g., Azure Blob Storage with encryption enabled). - DUpdate Terraform Practices:
Add .gitignore rules to prevent .tfstate files and other sensitive files from being committed to version control.
Implement sensitive = true on Terraform variables to avoid sensitive outputs being logged during plan/apply. - EImplement Code Scanning:
Use tools like GitHub Advanced Security, Checkov, or Terraform Compliance to detect sensitive information and security misconfigurations during code reviews. - FTrain the Team:
Educate developers on securely handling sensitive information. - GEnable Logging and Monitoring:
Enable Azure Monitor to track unauthorized access attempts. - HConduct a Post-Mortem:
Use Terraform logs to ignore secrets that might have been exposed. - ISecure Sensitive Data:
Store secrets directly in Terraform files with read-only access for all team members. - JUpdate Terraform Practices:
Avoid .gitignore rules, as sensitive files should be committed for transparency.
- AImmediate Actions:
13You are tasked with designing a CI/CD pipeline for deploying AI models into production. The pipeline must include steps for training the model, validating it, and deploying it into an Azure Machine Learning workspace.
Outline the high-level steps you would include in the pipeline- ASetup and Configuration:
Version Control: Store model code and configuration files in a Git repository.
Environment Configuration: Define environment requirements (e.g., Python dependencies, Azure ML SDK). - BModel Training:
Data Ingestion: Retrieve training data from a secure Azure Storage account or Azure Data Lake.
Train the Model: Execute training scripts in Azure Machine Learning pipelines or on compute clusters. - CModel Validation:
Validation Scripts: Run validation tests to ensure the model meets quality thresholds.
Bias and Explainability Testing: Evaluate the model for fairness and interpretability.
Model Registration: If the model passes validation, register it in the Azure ML Model Registry. - DBuild and Package:
Package the model and its dependencies into a Docker container for deployment.
Store the container image in Azure Container Registry (ACR). - EDeployment:
Staging Environment: Deploy the model to a staging endpoint in Azure ML Workspace for testing.
Smoke Tests: Run basic tests to verify the endpoint’s functionality. - FApproval and Monitoring:
Implement manual or automated approval gates before promoting the model to production.
Deploy the model to a production endpoint in Azure ML Workspace or AKS (Azure Kubernetes Service). - GPipeline Automation:
Automate the above steps using CI/CD tools like Azure DevOps Pipelines or GitHub Actions.
Use Terraform to provision required Azure resources for the pipeline. - HModel Training:
Train the model manually using a local Python environment and ignore Azure Machine Learning pipelines. - IBuild and Package:
Package the model without Docker and deploy it directly on a local machine instead of Azure Container Registry. - JApproval and Monitoring:
Skip approval gates and directly deploy the model to production without any testing.
- ASetup and Configuration:
14Which of the following best describes a key difference between DevOps and MLOps?
- AMLOps involves more stakeholders than DevOps, including Data Scientists, ML Engineers, and Data Engineers.
- BDevOps focuses solely on deploying ML models to production.
- CMLOps uses a linear development cycle, unlike DevOps.
- DDevOps includes continuous training of models.
15What is the primary focus of Continuous Integration (CI) in MLOps?
- ATesting and validating both code and data pipelines
- BAutomating deployments of ML models
- CManaging model versioning
- DMonitoring system throughput and resource utilization
16What is a major challenge in MLOps that differentiates it from traditional DevOps?
- AThe need to continuously validate and test code.
- BIterative cycles and experimentation on models.
- CThe absence of any need for CI/CD pipelines.
- DThe use of a monolithic codebase.
17In MLOps, continuous training is a part of the lifecycle to ensure that models adapt to changing data
- AVerdadero
- BFalso
18Which of the following are characteristics of MLOps? (Select all that apply)
- AVersioning of code only
- BDiverse team including Data Scientists, ML Engineers, and Data Engineers
- CIterative development cycle
- DFocus on monitoring CPU and throughput
- EContinuous training of models
19In MLOps, CD is responsible for deploying and managing [blank_start]__________,[blank_end] whereas in DevOps, CD is focused on deploying a single service or application.
- Amultiple pipelines
20Which of the following features in Azure is responsible for controlling access to resources based on roles?
- AAzure Firewall
- BAzure Policy
- CRole-Based Access Control (RBAC)
- DAzure Monitor
21Which of the following methods can be used to enhance data protection in Azure? (Select all that apply)
- AEncryption at rest
- BEncryption in transit
- CStoring passwords in plain text
- DAzure Key Vault
22Azure AD Multi-Factor Authentication (MFA) helps improve security by requiring users to:
- AProvide an additional authentication factor beyond username and password
- BChange their passwords every 30 days
- CAccess only from a corporate network
- DUse only biometric authentication
23What is the primary purpose of Public Key Infrastructure (PKI)?
- ATo store application logs securely
- BTo generate access control policies
- CTo issue, manage, and validate digital certificates
- DTo encrypt entire Azure VMs
24Which Azure service is recommended for managing certificates for PaaS applications?
- AAzure Security Center
- BAzure Key Vault
- CAzure DevOps
- DAzure Monitor
25Which of the following is true about private key certificates in Azure? (Select all that apply)
- AThey should be stored securely, such as in Azure Key Vault
- BThey are used for encrypting and decrypting messages in PKI
- CThey can be publicly shared without security risks
- DThey are issued by Azure Active Directory by default
26A virtual machine in Azure is unable to reach the internet. What is the FIRST thing you should check?
- ANetwork Security Group (NSG) outbound rules
- BAzure Virtual Network Peering
- CThe size of the VM
- DAzure Monitor Logs
27You need to resolve an issue where an Azure VM cannot reach a private DNS zone. Which actions should you take? (Select all that apply)
- ACheck if the VM is linked to the correct DNS resolver
- BRestart the VM to reset network settings
- CVerify the custom DNS settings in the virtual network
- DAssign a new public IP to the VM
28What tool can you use in Azure to troubleshoot network routing issues?
- AAzure Security Center
- BAzure Network Watcher
- CAzure AD Conditional Access
- DAzure Cost Management
29You suspect an NSG is blocking traffic to a VM. What should you do?
- ADelete the NSG and retry
- BDisable all security rules
- CUse Network Watcher to check NSG flow logs
- DReview NSG inbound and outbound rules
30What is the key difference between the "Owner" and "Contributor" roles in Azure RBAC?
- AContributor can manage access, while Owner cannot.
- BOwner has read-only permissions.
- COwner can assign roles, while Contributor cannot.
- DContributor has more privileges than Owner.
31Which role is best for a developer who needs to deploy and manage resources but does not need to assign permissions?
- AOwner
- BContributor
- CReader
- DUser Access Administrator
32A security team needs to audit resources without making changes. What role should be assigned?
- AOwner
- BContributor
- CReader
- DUser Access Administrator
33Which role allows a user to manage access permissions but not modify resources?
- AOwner
- BContributor
- CReader
- DUser Access Administrator
34 A junior engineer accidentally deleted a virtual machine. Which role assignment could have prevented this?
- AContributor
- BOwner
- CReader
- DVirtual Machine Contributor
35A user needs to monitor cost usage and invoices in Azure. What is the appropriate role?
- AOwner
- BBilling Reader
- CContributor
- DReader
36A team member needs to manage Azure storage but should not be able to change networking configurations. Which role should they have?
- ANetwork Contributor
- BOwner
- CStorage Account Contributor
- DVirtual Machine Contributor
37You need to grant access to manage only networking resources in an Azure subscription. Which role is appropriate?
- AContributor
- BReader
- CNetwork Contributor
- DOwner
38A company wants to follow the principle of least privilege. Which of the following are best practices? (Select all that apply)
- AAssign roles at the resource group level instead of subscription level.
- BGive all users Owner access for simplicity.
- CUse predefined roles before creating custom roles.
- DAssign multiple high-privilege roles to each user.
39An intern needs to review logs and reports but should not be able to modify anything. What is the best role?
- AContributor
- BOwner
- CSecurity Admin
- DReader
Información
- Descripción
- Test sobre ML Platform Engineer , creado por Junaid Khan el 27/01/2025.
- Autor
- Junaid Khan
- Idioma
- English
- Visitas
- 2
- Etiquetas
- Valoración
- Aún sin valoraciones Valora este test