Jon Schneider at SpringOne Platform 2017
summary
Project Micrometer is an instrumentation facade that records counters, gauges, timers, histograms and distribution summaries once and converts them for different backends, exports to Prometheus Atlas New Relic Datadog Influx, and auto-wires HTTP and scheduled method metrics. It adds coordinated omission compensation and a rolling max metric so GC pauses and outliers surface — demo shows average latency ~12 ms, 99th ~35 ms but rolling max spiking to ~350 ms; a Netflix service at 100k rps had max 15–20× worse about 1,000 times/sec.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 0:38Hospital analogy
- 2:47Importance of metrics
- 5:00Introducing Micrometer
- 6:19Alerting challenges
- 8:34Timer implementation
- 11:43Coordinated omission
- 12:45Measuring user experience
- 16:21Conclusion
related talks
36mMetrics - What's your code actually doing?
38mObservability 101 with Spring and Micrometer
47mMicrometer: New Insights into Your Spring Boot Application
44mWelcome to the Metrics
01h 08m#Micrometer Mastery Unleash Advanced #Observability in #JVM Apps
48mObservability of Your Application
53mUnmask hidden issues with Observability superpowers!
59mTo Production and Beyond: Observability for Modern Spring Aps