Stabilising eccentric systems
summary
Practical checklist for stabilising a messy system: run event storming and capability mapping to map product boundaries; audit docs, tests and CI/CD (one case shows a pipeline four years without a deploy); add structured logs, tuned alerts and R/E/D metrics (rate error duration) and track MTTD/MTTR from incidents; capture pain on post-its, group into security/commercial/maturity/architecture/KPIs/housekeeping and prioritise by urgency/impact/effort into Do/Plan/Backlog/Eliminate; iterate with small experiments and measurable targets.
AI-written summary · about
chapters (25)
- 0:00Introduction
- 2:10Agenda
- 2:30Preloved
- 3:36Language
- 5:19Domain
- 7:51Stakeholders
- 9:48Useful techniques
- 10:54Documentation
- 14:13Pipelines
- 16:28Observability
- 18:50Recent incidents
- 24:56Prioritization
- 31:49Workings
- 32:56Prioritize
- 33:26Burn
- 35:03Keeping it visible
- 36:18Making mistakes
- 36:39You cant fix the world
- 37:48Endings
- 39:23Building trust with teams
- 44:38Recommended resources
- 46:56Questions
- 47:13Comic Sands
- 48:43Sticky Notes
- 54:25Backlogs
related talks
26mBreaking Systems for Resilience
40mControlling Wildfires While Only Getting Singed.
26mFrom friction to function: Strengthening the product-engineering alliance
30mTaming the Chaos: Beyond the Quick Wins
32mClear the ring for Chaos Engineering
44mAvoiding the pitfalls of rebuilding software
46mMaster Software Architecture: From Simplicity to Complexity
33mTaming The Chaos: After The Quick Wins