Understanding What's Wrong With Your APIs - Before Your Boss Calls You
summary
Shows a working New Relic stack that stitches mobile front end to backend traces so a single mobile SDK captures device, OS, user id, session and a trace id that links to backend APIs. Dashboard demos include a funnel showing OTP drop-off, per-device filters (Xiaomi, Samsung), session-level traces, and alerts with incident correlation and ML-suggested root causes. Also shows runbooks attached to incidents, anomaly detection for spikes, and examples where OTP and empty backend responses break bookings.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:55API observability design
- 8:32Why holistic API observability
- 10:18User frustrations exercise
- 15:02Identifying what you need to know when APIs fail
- 27:28API dashboard demo
- 48:20Alerts and root cause analysis
- 55:00Summary and Q&A
- 59:37Closing remarks
related talks
33mBad API Observability
20mThe (Full)stack Trace: Understand Your App with Distributed Tracing
17mUnlocking Observability with React & Node.js
17m“How’s the App Doing?” Bringing Mobile Into Your Reliability Picture
25mFollow The (Full)stack Trace: Understand Your App with Distributed Tracing
26mObservability for Your Internal Developer Platform: Instrumenting Bac...
38mEffective App Monitoring in Production
19mOops my app performance sucks... what if observability could be a game changer?