The mainstreaming of containerization and microservices is raising a critical question of both developers and operators alike: how do we debug all this? Debugging has often been the most challenging and misunderstood aspect of application development and delivery, but that challenge grows ever more complex with modern, composed applications: while distributed systems can be highly resilient to certain kinds of failure, they’re exposed to new failure modes. In this talk, we will describe the debugging techniques we have employed, and like any good talk on debugging, this one will be laden with real stories from the trenches — and how those painful experiences resulted in better tools and better methodologies.