My Node.js process is on fire
summary
Black Friday traffic forces autoscaling to four times the servers and the bill spikes to 4× cost so the aim becomes a measurable goal: reproduce the spike, run cannon to measure latency and throughput, collect traces, and iterate. Use clinic doctor to reveal event‑loop delays and GC pressure and clinic flame to find CPU hotspots, fix the identified bottleneck, measure again, and repeat until the target latency and concurrency numbers are met.
AI-written summary · about
chapters (9)
- 0:00Handling performance issues
- 1:24The post-mortem process
- 3:05Fixing production bottlenecks
- 3:44Node.js concurrency and memory
- 5:16HTTP load testing
- 6:06Clinic Doctor analysis
- 8:10Clinic Flame profiling
- 9:26Deep dive into performance
- 10:48Avoiding over-optimization
related talks
25mMy node.js process is on fire
26mObserving Node.js: Using Metrics to Improve your Application Performance
46mNodeJS Scalibility Tips
01h 12mA New Way to Profile Async Activity in Node.js
21mCan you tell me if your Node app is healthy?
27mScaling NodeJS beyond the ordinary
29mReady, steady, crash!
50mDebugging Node.js in Production