We've configured our ULS Logs to be a maximum of 50GB, via Central Admin => Monitoring => Configure Diagnostic Logging. When clicking OK at that page, it will trim all logs on all servers to 50GB. However, over time it will grow to way more than 50GB, e.g. 200GB.
Only when we go back to this page and click OK, it will trim back to 50GB.
I can't find any timer job that is responsible for trimming the ULS Logs. Why isn't this working as expected?