"Scaling A/B testing on Netflix com with Node.js"

on YouTube ↗20m

" summary

Netflix uses Node.js to resolve conditional front end dependencies at build time and request time, avoiding prebuilding a combinatorial blowup — about 2.5 million packages from hundreds of concurrent A/B tests — by authoring annotations with NF-include, scanning with a build-time asset registry, and assembling packages via a runtime packager. Registries bundle file contents, deploy to CDN, and serve concatenated JS/CSS from memory in less than five milliseconds, and the system is in production use.

AI-written summary · about

" chapters (11)

1. 0:00Start
2. 0:31Alex Liu
3. 0:47Background
4. 1:38Multivariate Av Testing
5. 6:38Conditional Dependencies
6. 8:51Asset Registry
7. 10:37Packager
8. 10:48Api
9. 12:18Recap Using Visuals
10. 13:45Hash Attribute and File Size
11. 18:16Learnings

" related talks

• 22mObservable Node.js ApplicationsYunong Xiao · NodeConfEU 2015
• 31mKeynote Panel: Node.js at Massive ScaleY. Xiao, J. Chan, M. Hiskes, A. Slaton, K. Borchers · Node+JS Interactive 2018
• 27mBreaking the monolith with Node and Docker at NetflixTsubomi Imamura · JSConf Iceland 2016
• 26mGoing Faaster: Function as a Service at NetflixYunong Xiao · NodeConf EU 2018
• 29mBreathe Into the Node.jsMark Hinkle · Cloud Foundry Summit Silicon Valley 2017
• 31mNode.js that's hugely reliable, fast, and scalableMatthew Clark · NodeConf London 2016
• 31mGoing FaaSter: Function as a Service at NetflixYunong Xiao · Node+JS Interactive 2018
• 13mLessons learned in productionMatt Eernisse · NodeConf 2011