Hugo Kaczmarek (Product Manager) and Markus Hirt (Director Engineering) @Datadoghq talk about continuous application profiling in production, having better visibility in application environments, and managing performance budgets.
SHOW: 473
SHOW SPONSOR LINKS:
CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw
PodCTL Podcast is Back (Enterprise Kubernetes) - http://podctl.com
SHOW NOTES:
Topic 1 - Welcome to the show and welcome back to the show. Before we dig into today’s topic, can you give us a little bit of your background and how you became focused in this area?
Topic 2 - Production used to be a pretty sacred and untouched environment. Now we’re seeing the need to create chaos in production, test in production, and know more about how things actually work in production. Let’s talk about this concept of “profiling”.
Topic 3 - People often talk about having to “instrument their code” to be able to get feedback. How does that differ from profiling, and what unique things does an application need to be profiled in production?
Topic 4 - On the surface, this seems like it would add quite a bit of overhead to the applications. How do you manage to find the right balance of performance and visibility?
Topic 5 - Help us understand how profiling fits into the bigger picture of all the tools needs - Logging, APM, Observability, etc?
Topic 6 - Is there a profile of who needs profiling? Is this every application team, or just for specific applications (e.g. event-streaming, etc.)?
FEEDBACK?