It’s the middle of the night, and there’s much ado about nothing. Well, not quite about nothing - there’s definitely something happening: one of the services your team owns is crashing. And by crashing, I mean sometimes crashing. It’s not really that bad, but it’s bad enough for the floodgates to open and for the DevOps folks to lose their marbles and ping you relentlessly on Slack. Ping. Ping. Ping.

This talk is all about that feeling, and what we can do to make the whole situation suck less when it inevitably comes up again. We'll discuss the types of problems one can (and cannot) foresee, and offer technical and cultural steps developers can take to increase observability while on-call (even when a serious, hard to debug issue arises).