Python

Today in Python production.

- Logging had stopped working in subprocesses, for all loggers
- Looks like this had been going for a while (we did not notice, because no errors)
- Cause was APScheduler 3.10 -> 3.11 upgrade
- Still no idea what happened inside APScheduler patch that caused this

Oof, that sounds like a tough one. Silent failures are the worst. Any leads on what changed in APScheduler 3.11 that could have messed with logging? Maybe something with how it handles subprocess contexts or threading?