Server-side JavaScript the untold story
summary
A production e-commerce platform runs Rhino-based server-side JavaScript to render roughly 40 million pages/day, with a thin Java API layer, dynamic WSDL/Vistal compilation into Java, and an editor-only type-annotation hack for completion. Operational lessons drive hard constraints: disabled eval and quota limits on string, object and XML sizes to prevent memory storms, avoid SQL exposure, prefer copying data into isolated JS objects rather than writable wrappers, and cache compiled services to contain faulty or malicious custom code.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 2:33Server-side JavaScript in SaaS
- 5:46Rhino engine choice
- 8:56API capabilities
- 14:57SaaS constraints
- 19:11Implementation lessons
- 22:44Future plans and CommonJS
related talks
29mhome-grown server-side scripting
46mEmbracing Collaboration with JRuby and JavaScript
20mJavaScript 2 Enterprise Edition
22mJSConf 2015 Andrew Rota
37mRunning JS on the Server at Scale — What Can Go Wrong?
53mThe Unbearable Weight of Massive JavaScript
39mTowards Universal JavaScript
37mRunning JS on the Server at Scale — What Can Go Wrong