SRE practice map
An interactive prerequisite graph of about 190 reliability engineering practices: what to adopt, and what has to come first.
Open the mapWhy this site
The upstream r9y.dev reliability map is a strong catalog of engineering practices with a rare quality: it covers culture and organization alongside infrastructure. Its visual grid of maturity eras suggests an adoption order, yet the underlying data carries almost no dependency information: 16 explicit edges for 190 practices.
This project keeps the catalog and replaces the implied ordering with an explicit prerequisite graph. The map opens with the practices that require nothing else; expanding a practice reveals what it unlocks, so a team can trace a concrete adoption path instead of measuring itself against a maturity level.
The graph is data, not a drawing. Every deviation from upstream is marked - removed, added, split, reworded - dashed edges are drafts awaiting review, and the dataset invariants are enforced by tests: no cycles, no dangling references, no practice that unlocks nothing.
What to read
SRE foundations
- Site Reliability Engineering The book that defined the discipline: SLOs, error budgets, toil - the vocabulary this map is built on.
- The Site Reliability Workbook Worked examples for turning the first book into practice.
- Seeking SRE What the discipline looks like outside Google, told by many voices.
- Implementing Service Level Objectives The deepest treatment of the SLI, SLO and error-budget chain at the heart of this map.
Observability and operations
- Observability Engineering Why high-cardinality events beat dashboards of predefined metrics once systems get distributed.
- Distributed Systems Observability A compact map of logs, metrics and traces and their trade-offs.
- Practical Monitoring Monitoring as a practice, not a tool purchase; pairs well with the map's observability line.
Culture and incident response
- Blameless PostMortems and a Just Culture The post that made blameless retrospectives an industry norm.
- The Field Guide to Understanding 'Human Error' Why "human error" is a symptom of a system under pressure, never a diagnosis.
- How Complex Systems Fail Eighteen short theses; the density of insight per page is unmatched.
- PagerDuty Incident Response A complete, battle-tested incident command process you can adopt as is.
Engineering practice and delivery
- Accelerate The research behind the delivery practices on the map's development line.
- Release It! Stability patterns - timeouts, bulkheads, circuit breakers - that many map nodes assume.
- The DevOps Handbook The flow, feedback and learning loops behind the cultural practices.
- Chaos Engineering Turning failure injection from a stunt into a discipline with hypotheses and blast-radius control.