Ask your logs
summary
Shows that turning Rails logs into structured JSON lets teams query them as data: use lograge or semantic_logger with a JSON formatter, enrich lines with user_id, revision and env, and correlate via a correlation ID or Sidekiq middleware adding request_id. Example outcomes: a dashboard in Kibana spots 800,000 enqueued jobs, filters reveal 20 requests by a user, seven legacy API users out of 109, and a feature-flag causing DB latency in us-east. Warns about performance cost and leaking sensitive params (Facebook plain‑text password incident).
AI-written summary · about
chapters (19)
- 0:00Introduction and story
- 2:30About the speaker
- 3:19Talk agenda overview
- 4:02Default Rails logs explained
- 5:48Questions logs should answer
- 7:52Problems with current logs
- 8:14Consolidating multi-line logs
- 9:41Structuring logs into JSON
- 11:03Enriching logs with domain data
- 13:58Correlating logs between systems
- 16:28Analyzing logs as data
- 17:30Querying user-specific requests
- 18:36Investigating job spikes
- 21:04Tracking Legacy API usage
- 23:14Cross-stack performance analysis
- 24:39Benefits of structured logging
- 25:00Historical context of structured logs
- 26:23Call to adopt structured logging
- 27:24Q&A and closing
related talks
28mAsk your logs
28mAsk your logs
23mSquash Production Defects Quickly - The Power of Structured Logging in Rails
30mLogging Rethought 2: The Actions of Frank Taylor Jr.
29mAre logs a software engineer's best friend? Yes - follow these best practices
42mMake your logs work for you
43mThe Dangers of Being Overly-Paranoid
01h 00mApplication Logging & Logstash