Devops@Scale
summary
Starts at three engineers and scales to ~100, moving through clearly named stages: fire brigade, smoke alarm installers, then a centralized DevOps enablement team and finally proactive improvement. Uses GitHub, Travis CI, AWS Beanstalk, CloudWatch and Pingdom for ops, JIRA/Confluence for incidents. Recovers from a global NPM outage (left-pad incident) and a GitLab six‑hour data loss by publishing transparent postmortems. Builds a load generator to hit a billion spans/min for elastic performance testing and measures petabytes and billions of records to decide priorities.
AI-written summary · about
chapters (14)
- 0:00Start
- 11:30Reactive Improvement Stage
- 12:28Exploratory Testing
- 14:12Pingdom
- 16:59Incident Report
- 18:58Document the Root Cause
- 19:28Document Action Items
- 21:11Gitlab
- 31:05Non-Functional Backlog
- 35:05Proactive Improvement
- 36:09Non-Functional Definition of Done
- 41:25Summary
- 43:45Questions
- 44:21How Do You Onboard New Team
