Microsoft Certified: Azure AI Engineer AssociatePlan and manage an Azure AI solutionMedium
A research institution is developing an Azure AI solution to analyze large datasets from scientific experiments. The solution uses Azure Machine Learning compute instances for training and Azure Cognitive Search for indexing and querying results. To control costs, the institution wants to automatically shut down compute resources when they are not actively being used, especially outside of working hours. Which Azure Cost Management feature or practice is most relevant for this scenario?
- AAutoscaling and Scheduled Shutdowns
- BAzure Advisor recommendations
- CAzure Reservations
- DAzure Budgets and Alerts
Show answer & explanationAnswer & explanation
Correct answer: A. Autoscaling and Scheduled Shutdowns
Autoscaling ensures resources are only active when needed, and scheduled shutdowns allow for automatically turning off compute resources during idle periods (e.g., nights or weekends), directly reducing costs for non-continuous workloads.
Why the other options are wrong
- B. Azure Advisor provides recommendations for cost optimization, but it's a recommendation engine, not an active management tool for shutdowns.
- C. Azure Reservations provide discounted rates for committed usage but don't address automatic shutdown of idle resources.
- D. Azure Budgets and Alerts help monitor and notify about spending, but they don't actively manage or optimize resource usage.
Scheduled Shutdowns & Autoscaling
Azure features that allow for automatically stopping compute resources during specified idle periods or scaling them up/down based on demand, respectively, to optimize costs.
- Reduces costs by paying only for active compute time.
- Scheduled shutdowns are ideal for development/test environments or predictable idle times.
- Autoscaling ensures performance while minimizing over-provisioning.
Memory trick: Scheduled Shutdowns are like 'turning off the lights' when no one's home.