Node.js at UBER
summary
A system for turning Node prototypes into production services with concrete pieces: a statsd proxy that needed to run on 50 machines, logging piped to Kafka and Sentry with automatic blame emails, health checks tied to Nagios and PagerDuty, and timeouts on every external HTTP call. A scaffolded CI tool called Potter generates repos that insist on 100% code coverage, wires in Graphite dashboards, Puppet/EC2 staging, and a control port for profiles and heap dumps; unhandled exceptions abort the process.
AI-written summary · about
chapters (6)
- 0:00Start
- 3:14Logging
- 4:18Error Reporting
- 6:46Configuration and Managing Configuration
- 9:14Dealing with Uncle Exceptions
- 13:03Local Workflow
related talks
30mJavascript @ Uber
20mBeing Uber with Node
30mNodeJS: Patterns and Tools to Ship to Production with Confidence
40mDeploying node.js at scale
31mNode.js that's hugely reliable, fast, and scalable
17mNodejs microservices. From Zero to Hero.
22mObservable Node.js Applications
27mProgressive Microservices with NodeJS