Log Rotation

* Scheduling Guide

Log rotation schedules prevent disk exhaustion on servers that produce continuous application, access, and error logs. The rotation frequency depends on log volume — a busy web server may need hourly rotation to keep files under 100 MB, while a low-traffic internal service can rotate daily. Standard Unix logrotate is triggered by cron (typically 0 0 * * *) and handles compression, renaming, and retention in a single pass. For application-level log management, the schedule should compress and archive logs to cold storage after a defined retention period and delete local copies once the upload is confirmed. Always send a signal to the application process after rotating so it reopens the log file handle — otherwise the process continues writing to the renamed (or deleted) file.