Bryan Cantrill interviewed at Velocity 2011
summary
Built a DTrace-based visualization for Node.js during the 2010 Node Knockout, instrumenting a distributed, dynamic, highly virtualized deployment and tying traces back to V8-based event loops. Uses live DTrace probes to map async callpaths, surface contention and latency, and correlate events across nodes. Finds Node’s enforced asynchronous model often yields high throughput without heavy manual tuning; one system replaces 10,000-line C with 400-line Node and runs in production. Recommends data-heavy tests under load and cross-node trace correlation.
AI-written summary · about
chapters (6)
- 0:00Introduction to Node.js
- 1:02A technical revolution
- 1:51Real-world applications
- 2:46Performance and architecture
- 3:40The future of the web
- 4:58Getting started with Node
related talks
22mExamining Observability in Node.js
25mNode js applications diagnostics under the hood
28mTracing Node. The Open Source Edition
08mComprehensive Observability via Distributed Tracing on Node.js8
01h 04mNode.js performance analytics and best practices
49mA Journey into Node.js Internals
57mTroubleshooting Node.js Applications Under the Hood
50mDebugging Node.js in Production