Under the right circumstances, error logs are magical. They can shed light on how a system is behaving - or misbehaving; and, they can provide a path forward in an emergency situation. But, logs do come at a cost. Not only is there a real-world dollars-and-cents aspect to aggregating and retaining logs, having too many logs can end-up reducing the signal-to-noise ratio which can make it harder to debug a failing application.
This week on the show, the crew talks about different categories of logging, why some logging is required for regulatory purposes, how you can replace certain logs with telemetry, metrics, and tracing. And, why you should absolutely never ever email yourself errors... except for when you should; because it's easy; and because we never learn from our own mistakes.
All that and more on this week's show:
... featuring these beautiful, beautiful people:
- Adam Tuttle → Website, Twitter, LinkedIn
- Carol Hamilton → Twitter, LinkedIn
- Tim Cunningham → Twitter, LinkedIn
- Ben Nadel (that's me) → Website, Twitter, LinkedIn
With audio editing and engineering by ZCross Media.
For the full show notes and links, visit the episode page. And, be sure to follow the show and come chat with us on Discord! Our website is workingcode.dev and we're @WorkingCodePod on Twitter and Instagram. New episodes drop weekly on Wednesday.