What SRE Could Be: Systems Reliability Engineering
summary
Argument: SRE is systems thinking applied to software and sociotechnical systems, not just SLOs or pager duty. Demonstrated tools include hierarchical task analysis, cognitive work analysis, EAST-BL and causal loop diagrams used to map goals, links and failure modes. Concrete examples: delaying aggressive re‑replication to reduce churn, cellular architecture to limit blast radius, initializing counters to expose rollovers, and a student exercise that surfaces 21 distinct Death Star faults. These methods scale as focused practices (a few percent of time) and produce actionable interventions.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 3:00SRE’s identity crisis
- 6:54DevOps vs SRE practices
- 9:21Standardizable vs deep engineering
- 11:49SRE as systems interventions
- 14:13Effective systems interventions
- 19:13Social systems and SRE
- 22:05Systems thinking importance
- 25:14Complex systems challenges
- 27:00From hands-on to mission control
- 29:07Systems thinking Death Star
- 31:11Informal systems thinking in SRE
- 32:14Cognitive work analysis
- 34:22Hierarchical task analysis
- 37:07Forgotten system control logic
- 38:13East BL broken links
- 39:36Causal loop diagrams
- 42:03Imperfect models in engineering
- 43:47Hazards constraints controls
- 45:34Beyond data-driven reasoning
- 47:03Adopting systems thinking
- 47:45SRE book as systems zoo
- 48:27Call to systematize SRE
related talks
35mCare and Feeding of SRE
25mSRE Guiding Principles & Transformational Framework
24mBeing Reasonable about SRE
45mClearing the Way for SRE in the Enterprise
43mThe SRE I Aspire to Be
35mThe Best SREs Seem to Be the Ones without an SRE Title—And What We Can Do
26mFrom Nothing to SRE: Practical Guidance on Implementing SRE in Smaller Organisations
41mAn Introduction to SRE at Google (with DevOps Focus)