Lessons in this course (9)

9. Cost optimisation for Azure databases

IntermediateMediumLesson 9 of 9
  • Right-size from metrics (many databases are over-provisioned); scale down at night with automation for non-production.
  • Serverless for dev/test and intermittent workloads (auto-pause).
  • Elastic pools for many small databases with uneven usage.
  • Reserved capacity (1 or 3 years) and Azure Hybrid Benefit for steady production workloads.
  • Read replicas for reporting instead of scaling the primary.
  • Cosmos DB: choose a good partition key (fewer RUs), autoscale or serverless, index only needed paths, use point reads.
  • Delete unused databases/backups; set LTR policies deliberately; monitor with Cost Management budgets and alerts.