Question 1: What are the main cost drivers for data engineering?

Answer: Compute (Synapse DWU, Databricks DBU, ADF activities), Storage (ADLS tiers), Network (data transfer), and Licenses (Azure Hybrid Benefit).

Question 2: How do you estimate costs for a new project?

Answer: Use Azure Pricing Calculator, TCO Calculator, estimate compute hours, storage GB, data transfer, and add 20-30% buffer.

Question 3: What is the benefit of Reserved Capacity?

Answer: 30-50% savings over pay-as-you-go. 1-year or 3-year commitment. Use for stable, predictable workloads.

Question 4: How do you optimize Synapse costs?

Answer: Auto-pause for non-24/7 workloads, right-size DWU, reserved capacity, and use serverless for ad-hoc queries.

Question 5: What is the difference between pay-as-you-go and reserved?

Answer: Pay-as-you-go: No commitment, higher cost. Reserved: 1-3 year commitment, 30-50% savings. Use reserved for stable workloads.

Question 6: How do you optimize storage costs?

Answer: Lifecycle management (Hot→Cool→Cold→Archive), compression, partition pruning, and avoiding small files.

Question 7: What is the benefit of auto-pause?

Answer: Eliminates compute costs when not in use. Resume automatically when queries arrive. Saves 50-70% for non-24/7 workloads.

Question 8: How do you monitor costs?

Answer: Azure Cost Management, budgets, alerts, cost analysis by service/resource, and chargeback reports.

Question 9: What is the cost impact of geo-redundancy?

Answer: Doubles storage costs (primary + secondary). Worth it for critical data (downtime cost often exceeds storage cost).

Question 10: How do you optimize data transfer costs?

Answer: Same-region deployment, Private Endpoints, compression, batch transfers, and avoiding unnecessary cross-region transfers.

Question 11: What is Azure Hybrid Benefit?

Answer: Use existing Windows/SQL licenses for Azure savings (up to 40%). Apply to VMs, SQL Database, and Synapse.

Question 12: How do you handle cost overruns?

Answer: Set up budget alerts, implement resource locks, use Azure Policy for guardrails, and review cost trends regularly.

Question 13: What is the benefit of Dev/Test pricing?

Answer: 60-90% discount for non-production workloads. Use for development, testing, and staging environments.

Question 14: How do you optimize Databricks costs?

Answer: Auto-scaling clusters, spot instances, right-sizing, job clusters (not interactive), and reserved DBU capacity.

Question 15: What is the cost difference between Serverless and Dedicated Synapse?

Answer: Serverless: Pay-per-TB-scanned (variable). Dedicated: DWU-hour (reserved). Use serverless for exploration; dedicated for production.

Question 16: How do you implement chargeback?

Answer: Tag resources with team/project, use Azure Cost Management for allocation, and create chargeback reports.

Question 17: What is the benefit of Azure Advisor?

Answer: Recommends cost optimizations, performance improvements, security enhancements, and reliability improvements.

Question 18: How do you handle unexpected cost spikes?

Answer: Set up alerts at 80%/100% thresholds, implement auto-shutdown, use Azure Policy for limits, and investigate root cause.

Question 19: What is the cost of data lake storage tiers?

Answer: Hot: $0.018/GB/mo. Cool:$0.01/GB/mo. Cold: $0.0045/GB/mo. Archive:$0.001/GB/mo. Use lifecycle management for automatic tiering.

Question 20: How do you optimize ADF costs?

Answer: Use appropriate integration runtime, minimize activity runs, use Mapping Data Flows only when needed, and implement efficient scheduling.

Question 21: What is the benefit of Azure Cost Management?

Answer: Cost analysis, budgets, alerts, recommendations, and export capabilities for cost management.

Question 22: How do you handle multi-subscription cost management?

Answer: Use management groups, Azure Cost Management across subscriptions, and consolidated billing.

Question 23: What is the cost of network egress?

Answer: Varies by region and volume. First 5GB/month free. Use Private Endpoints and same-region deployment to minimize costs.

Question 24: How do you optimize Cosmos DB costs?

Answer: Right-size RU/s, use autoscale, choose appropriate consistency level, and optimize partition key for even distribution.

Question 25: What is the future of cost optimization?

Answer: AI-driven recommendations, automated optimization, serverless-first architectures, and FinOps practices.