As a software engineer responsible for building and operating various async and background workflows, I get a lot of inquiries like "I see cause X, why wasn't t

As a software engineer responsible for building and operating various async and background workflows, I get a lot of inquiries like "I see cause X, why wasn't there side effect Y?" and I go and poke around the data to figure out what happened. Sometimes the cause didn't really happen, sometimes there are other criteria that the stakeholder wasn't thinking about that weren't met, sometimes the cause really happened but the data is missing or malformed upstream, and very occasionally my system errored out and dropped the ball. These are all deeply mysterious head-scratching issues if you don't have ready access to detailed logs. More mundanely, if I get paged about an error rate, the first thing I'm going to do is look at the logs to see what the errors are.