The Journey of Reliability: Why bother at all?
summary
Claims a practical observability maturity model tied to concrete stages and tooling rather than buzzwords, with a proven build of a time‑series data warehouse at billion cardinality that powers Disney Hotstar and a Cassandra data stack, and shows what actually works. It maps problems to stages — instrumentation → emission → storage → query → dashboard — says use logs metrics and traces together, notes tracing is required for service‑dependency questions, warns costs rise with more visibility, and prescribes the incident recipe identify impact then notify stakeholders don't fix so debugging comes last.
AI-written summary · about
chapters (12)
- 0:00Why observability matters
- 2:06Debugging with the razor
- 6:02What is observability
- 7:05Four expectations from systems
- 8:52Logs metrics traces all three
- 10:38When systems fail lifecycle
- 13:50Start with the right data
- 15:57Tool classification by phase
- 17:23Evolving observability questions
- 20:28Translucency vs transparency
- 22:55Pick the sweet spot
- 24:20Outcomes before tools
related talks
37mBetter Reliability Through Observability and Experimentation
22mReliability Reviews in the Wild: Using Data to Drive Production Health
35mIs My Thing Reliable? Adventures in Driving Consistency of...
21mBeyond Observability - Aligning Technology Performance to Business Outcomes
29mThe Curious SRE: Cultivating a Mindset for System Resilience
22mLeading software teams with context (visibility)
20mBuilding a Culture of Reliability
33mThe Value of Reliability