Why should you care?

Black Friday traffic should be a slider — not a weekend hardware delivery.

See it live — copy this example

Run in SQL Server Management Studio (SSMS) or Azure Data Studio.

-- Measure pressure before scaling
SELECT * FROM sys.dm_db_resource_stats; -- Azure SQL DB
-- Or on MI/VM: sys.dm_os_performance_counters / wait stats
-- Then scale tier in portal/CLI and retest
SELECT DATABASEPROPERTYEX(DB_NAME(), 'ServiceObjective') AS ServiceObjective;

What happened?

• dm_db_resource_stats shows CPU/IO/log usage over time on Azure SQL DB.
• Scale only after evidence; retest critical queries after scaling.

Practice next

1. Capture resource stats during a load test.
2. Identify CPU vs IO vs log bottlenecks.
3. Scale up one tier and remeasure.
4. Create a read replica and point reports there.
5. Compare Hyperscale vs General Purpose for growth.

Remember

Measure, then scale. Match bottleneck to lever. Retest after changes.

Festival scale-up

DataVerse raises vCores for Diwali week, then scales down.

Outcome: Performance holds; bill returns to normal after.