Reliability Scoring: A 3-Part Formula for Promoting Reliable Code
summary
A 3‑part formula scores releases 0–100 by penalizing new errors, rate increases and slowdowns, aggregates penalties into a single reliability score, and plugs into open source reliability scoring visualized in Grafana and CI/CD plugins. The system detects and classifies events, blocks low scores from promotion, and the team reports it cuts sev1 incidents by 75% after enforcing quality gates and build blocking.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:21Reliability in software
- 3:09Speed versus stability
- 5:59Improving reliability principles
- 7:25Quantifiable quality metrics
- 9:28Measuring code quality
- 11:15Reliability score concept
- 13:01Error detection to prioritization
- 15:27Classifying errors
- 17:14Handling errors and slowdowns
- 18:383-part reliability formula
- 20:45Open sourcing reliability score
- 22:09Results on production outages
- 22:51Quality gates in CI/CD
- 23:14Q&A
related talks
13mBeyond observability: Using reliability scores to drive results
34mBuild Scoring: A Formula for Blocking SEV1s from Production
47mWhy Is My App SLOw? Defining Reliability in Platform Engineering
27mContinuous Reliability. How?
37mRelia...bility?
18mMeasuring Reliability Culture to Optimize Tradeoffs: Perspectives from an...
24mIntroducing the Reliability Map – r9y.dev
17mScaling End to End Reliability Tracking Across Large Scale, Multiplexed Products and Services