Beyond Traces - The Insights in Trace Aggregates

Explore the power of trace aggregates in distributed systems through this 33-minute conference talk from Strange Loop. Discover how aggregating traces can reveal profound insights beyond traditional tracing practices, including resource contention sources, latency anomalies in service infrastructure, and correlations with high-cardinality system characteristics. Learn about novel applications of aggregated traces for performance improvements, the importance of standardized tracing output formats, and cloud-native service mesh integration. Gain valuable knowledge on distributed system performance analysis, NoSQL, privacy & security, and how to provide developers with meaningful performance signals.

Intro
Quick review
Trace aggregate analysis
Correlation based on statistical analysis
Latency anomalies in context of service infrastructure
Resource contention analysis
Example: mutex contention terminology
Backup Slide for Murphy's Law