How to collect application metrics if it needs to be done yesterday
summary
Claim: don’t blindly ship a black‑box monitor — prefer targeted metrics and manual instrumentation so ingest costs stay low. Examples: use New Relic metrics with buckets per subsystem, split front-end vs API and user‑type dimensions, measure Web Vitals plus SPA route changes and Android crash/ANR rates. Expect a legacy app to take ~half a year and 2–3 people for full coverage; instead prioritize perceived performance, autoscale thread pools and add cloud balancer automation to cut on‑call noise.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:27Metrics for users and business
- 2:53Open source vs third-party tools
- 3:55Detecting problems early
- 6:23Perceived performance
- 7:47Dimensional metrics for apps
- 12:19Front-end monitoring with Web Vitals
- 15:50SPA metrics challenges
- 16:32Mobile app metrics
- 17:15Thread pools impact
- 21:09Alert fatigue and alerting
- 23:59Automation for on-call
- 25:02Post-mortems and sharing
- 26:26New instrumentation practices
- 27:08Dashboard design
- 28:11Team communication
- 28:54Observability process
related talks
34mMonitoring for devs. The necessity or self-indulgence
38mEffective App Monitoring in Production
05mA 5 Minute Checklist for Application Monitoring
01h 04mDevOps Without Measurement is a Fail
37mMonitoring the Health of Your App
14mBeyond the Basics: Performance Monitoring and User Experience for Mobile App Growth
17m“How’s the App Doing?” Bringing Mobile Into Your Reliability Picture
01h 03mMeasure All The Things with App Metrics