Achieving DORA outcomes by building reliability
summary
Argues that high reliability comes from software and platform design not magic infrastructure, recommends measuring the four DORA metrics (use Four Keys for implementation), codifying deployment archetypes tied to failure domains (hot‑cold two‑zone gives ~99.98% SLO ceiling), practice gradual change and mitigations (drain/spill/rollback), and build an internal developer platform as a product. Describes an MVP lab called RAP that runs Hipster Shop across multi clusters, costs about $50/day to run, and survives a cluster outage in live demo.
AI-written summary · about
chapters (11)
- 0:00Introduction and hook
- 3:29Pyramids of reliability
- 7:03Sublinear scaling and automation
- 13:20DORA metrics and platform building
- 17:33Application archetypes and failure domains
- 20:41Lab demo and reliable app platform
- 22:45Production rollout strategies
- 23:48Parting shots and key takeaways
- 24:55Q&A on DORA metrics
- 27:00Adoption of application archetypes
- 28:24Funding reliability initiatives
related talks
22mDORA Metrics: The Key to DevOps Excellence
25mTroubleshooting at Scale: What DORA Metrics Miss
23mOptimizing Delivery with DORA Metrics: How We Prepared and Where to Start
30mIntroducing DORA Core: Durable research insights for technical practitioners
28mAccelerating Success: How to Optimize Value Delivery with DORA
39mThe Secrets of High Performing Technology Organizations
19mAccelerating Code Quality with DORA Metrics
46mA Decade with DORA