Finding the Needle in the Stack Trace: APM Logs-in-Context
summary
New Relic ships an APM logs-in-context feature that injects span and trace IDs into logs, forwards logs directly from the Ruby agent (v8.7+), and links log events to transactions so logs appear beside stack traces and distributed traces in the UI. Instrumentation hooks use Ruby logger format_message via module pre-pending or alias chaining. Sampling kicks in at 10,000 logs per 60s (configurable up to 100,000). Early users like New York Public Library and Chegg report reduced toil and lower log volume.
AI-written summary · about
chapters (8)
- 0:00New Relic APM logs-in-context
- 3:10Log data collection challenges
- 5:35APM logs-in-context solution
- 5:55Ruby agent instrumentation
- 15:53Demo of logs-in-context in New Relic
- 17:42Customer success with logs-in-context
- 20:11Future of logs-in-context
- 21:37Q&A on agent upgrades
related talks
38mApplication Performances Monitoring with Elastic APM and the ELK Stack
52mNot on Our Watch: an Introduction to Application Performance Monitoring
01h 04mTracing distributed services: experiences with implementing APM for the JVM
41mFinding the Needle in the Haystack or Troubleshooting Distributed Systems
30mRails Performance monitoring: A primer for developers
20mThe (Full)stack Trace: Understand Your App with Distributed Tracing
39mPanel Discussion: Application Monitoring
32mWhat’s my App *Really* Doing in Production?