What was your Python up to at 3am?
|Countdown link||Open timer|
If a customer was hopping mad overnight, but your monitoring dashboard shows its usual green line at 99.5% superb, you might expect to spend your morning wading through logs… but what if you could find out what was happening in a few minutes? Better yet, what if you knew before they did?
For years, I've barely touched logs: I've instead used structured events and tracing to find the source of someone's trouble and list everyone else affected. It used to require custom code, but you can do it with open source: OpenTelemetry is an ambitious attempt to standardise application tracing across all languages, frameworks, packages, and deployment environments.
In this talk you'll learn how to measure your Python's behaviour with OpenTelemetry, how to solve production mysteries, and how to monitor service level indicators for advance warning of trouble.
I regret I lack the time to complete my biography right now.