Investigating production issues in a microservice architecture can make you feel like Sherlock Holmes, searching through evidence and gathering sources to recreate the scene of a crime. Often, this investigation involves digging into multiple log stores and dashboards to piece together an understanding of an issue. This is costly for engineering teams and customers – time spent sifting through clues only further delays the resolution. Distributed tracing can help. With distributed tracing, we see a request’s path through a complex system. But is that enough? At Netflix, we have taken distributed tracing, added in log correlation (with high-quality, detailed logs) and layered analysis on top. Not only does this reduce time for engineers to understand the root cause of an issue, we’ve provided this tool to customer operations, empowering them to understand the root cause and escalate with clarity. You’ll leave this talk with an understanding of distributed tracing and how to supplement traces with logs. You’ll see examples of how to shape your logs to clarify the business logic underneath a microservice’s response, and you’ll understand how tracing is the lynchpin to the type of detailed insights that will cut down on the cost of your team’s operational burden.