How to improve node.js perfomance with CPU profiling?

A little study case to understund this concept in practice

Hey everyone 👋

You already thought how a node.js system works at Black Friday or Christmas weeks? And most important, how to improve performance to deal with huge loads of requests?

To answer this, I start studying CPU profiling and JS performance at all (V8 relationship througth node.js), and the results are interesting.

What i have used?

• The standard of the market for evaluate this is Open telemetry, but for this case, I used the native node.js module called Inspector (node.js version 22+)
• For response time and load test I used Autocannon

What is the code?

• I have a simple node.js web server with 2 routes, one using lodash to manipulate data (picture 1) and other using the native array methods of JavaScript (picture 2)
• After getting data with inspector, I put this inside a file with .cpuprofile extension and upload him to https://discoveryjs.github.io/cpupro/

Results:

Code: https://github.com/henriquesss/cpu-profiling-nodejs

Used resources

• this video: https://www.youtube.com/watch?v=3F1uBQVWNfE
• this docs(for basement): https://nodejs.org/docs/latest-v22.x/api/inspector.html

Bye 👋