Measuring the Health of Your CNCF Project: Going Beyond Stars and Forks
Talkat KubeCon + CloudNativeCon North America 2021 · on Thursday, 2021-10-14· at 14:30· 411 Theater + Online·
summary
Stars and forks don’t prove health; measurable signals do. The talk shows using devstats and CHAOSS-derived metrics to track time-to-first-response, median issue close time (often a few days in healthy projects), contributor counts and churn, release cadence, and organizational diversity. Examples: Prometheus and Envoy show predictable holiday dips (Dec, Jul–Aug) and post-conference spikes; spikes in opened issues correlate with temporary rises in median close time. Recommendation: pick a few metrics, use year and monthly windows, and drill into repos or orgs to act on specific problems.
AI-written summary · about
chapters (12)
- 0:00Intro
- 3:18Focus on areas for improvement
- 3:58Every CNCF project is different
- 6:47Keeping up with contributions helps with retention
- 11:42Do people want to stick around?
- 15:50Lower risk many contributors and organizational diversity
- 17:39Are contributions increasing, steady, or decreasing?
- 19:26Regular releases and security updates
- 22:11Have a regular Cadence for evaluating project health.
- 24:22Whole project or repository group
- 25:02Linux Foundation's TODO Group
- 25:47Project health metrics require interpretation and verification
related talks
43mBuilding and Supporting Open Source Communities Through Metrics
44mIs this Open Source Project Healthy or Lifeless?
43mIf You Can’t Measure It, You Can’t Improve It
44mWhy Is My Community In Danger?
26mBuilding & Supporting Open Source Communities Through Metrics
36mPanel Discussion: OSS Viability and Project...
35mThe Benefits and Pitfalls of OSS Project Metrics: Measuring Health and Risk in Open Source Communities
46mMeasuring Community Health