Demystifying Azure App Service's Performance and Scaling
summary
Explains Azure App Service performance and scaling behavior: App Service scales only horizontally (changing SKU size causes a reboot/downtime), and clusters of smaller P1 instances can outperform a single larger P3. Presents performance test results (claims up to 30,000 requests/sec on small instances), discusses cost (pricing per App Service Plan and savings by hosting many apps together), and shows CLI features for per-app instance caps, scheduled scaling, and cooldowns to avoid flapping.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 2:33Azure App Service overview
- 5:05App Service plans and pricing
- 7:49Choosing App Service tier
- 13:30Right size for app
- 15:00Horizontal scaling limits
- 19:03High CPU not a problem
- 20:28Stacking explained
- 22:38Scaling basics
- 24:03Advanced scaling metrics
- 26:47Scaling rules setup
- 28:11Performance monitoring signals
- 30:42Prevent scaling flapping
- 31:24Avoid excessive alerts
- 33:17Interpreting metrics
- 34:43Reactive vs proactive scaling
- 35:08High availability setup
- 36:14Why autoscale isn't first
- 37:31Performance testing importance
related talks
01h 01mSupercharge your App Service to a Global scale
43mUsing AI with App Service to deploy differentiated web apps and APIs
52mSpring Apps at Scale from 0 to 5B to 100B+ Requests
43mDatabase capabilities you need for your next awesome app Q&A
15mWell-Architected Framework: Performance Efficiency
21mMusic Matched to Different Ways to Scale in K8s, What Could Go Wrong?
44mIntroduction to application scaling
58mDay 2 Session 1