Angular Performance in the Enterprise
summary
A demo dashboard is profiled with the Chrome DevTools profiler after running an ng build --prod with mangling disabled and in incognito; the profiler surfaces a leaked setInterval that triggers redundant change detection, and an aggregate getter that consumes over 50% of runtime. The expensive getter is moved into a web worker via ngg web worker and a small service that returns "calculating" then the numeric result, which removes the blocking work. Refactoring parent components to onPush and adding virtual scrolling cuts rendered components; enabling the V8 timeline shows JIT compile spikes of 5x–10x.
AI-written summary · about
chapters (8)
- 0:00Intro
- 1:13Optimizing an Angular Application
- 2:30Profile an Application
- 5:59Flame Graphs
- 8:25Flame Charts
- 9:38What triggers change detection
- 11:08Long calls
- 16:10Rendering
related talks
31mAngular Performance: Your App at the Speed of Light
19mAngular Performance Patterns
43mAngular Performance Workshop
26mProfiling Angular apps like a Shark
12mAngular performance pitfalls and how to avoid them
34mHigh-Performance Applications with Angular
35mPerformance at Scale
23mPerformance optimizations in Angular