Lightning Talk: Observability Debt: When Telemetry Stops Telling the Truth
summary
Claims that observability debt causes telemetry meaning to drift so dashboards and alerts lie: a simple API + queue + async workers model shows P99 latency appears to drop 64% while users actually wait >2s, 15% of jobs silently fail, and about 40% of incidents start with actively misleading signals. Detection takes roughly 3× longer, typical timelines hit 45–90–120 minutes. Fixes are concrete and non-disruptive: run telemetry archaeology after architectural changes, instrument full user journey with correlation IDs and versioned SLOs, then audit one dashboard.
AI-written summary · about
related talks
13mLightning Talk: Observability-First DevSecOps: Building Resilient Multi-Cloud P...
10m⚡ Lightning Talk: Unlocking Customer-Centric Observability: A Case Study of OpenTelemetry To Reduce MTTD < 3 Mins
38mThe Ticking Time Bomb of Observability Expectations
39mThe Ticking Timebomb of Observability Expectations
42mObservability Is Not Analytics!
24mObservability Through Users' Eyes and How To Fix the Knowledge Disc...
17mLightning Talk: Avoiding Chaotic Dysfunction: The ‘Don’t Do This’ Guide T...
25mNo observability without theory