Plex can keep drives awake without an active viewer because scheduled tasks, metadata work, database access, or another service may still touch the storage.
The first goal is to identify which process opens the device and when. Do not immediately extend spin-down timers or disable every Plex feature; capture a few wake events and correlate them with scheduled tasks, logs, scans, and other containers. A repeatable timestamp usually narrows the cause faster than changing power settings blindly.
Correlate Wake Times With Scheduled Plex Work
Idle to the viewer is not necessarily idle to the server. Scans, metadata refreshes, preview generation, maintenance, and analysis can run on a schedule after streaming stops.
idle Plex disk activity can coincide with scheduled or background tasks even when nobody is streaming.
Record the exact disk wake timestamps for a day and compare them with Plex scheduled-task windows and logs. If wake-ups align with one recurring task, move or disable only that task before changing drive power policy.
Separate Media-Drive Reads From App-Data Writes
Plex can update its database or metadata even when it is not reading a movie file. If app data and media share one spinning disk, tiny state operations can wake the same device that stores large files.
separating app data from bulk media lets metadata I/O and large media reads use different storage paths.
If possible, monitor the media path and Plex app-data path separately and identify which directory is touched first. If small app-data writes are waking the media disk because both share one volume, consider separating those roles before disabling useful Plex maintenance. A media-center layout with separate application state and bulk media can let Plex update fast state without waking the media disk for every small operation.
Check Other Containers on the Same Mount
Downloaders, indexers, backup tools, antivirus, and file watchers can wake a disk while Plex happens to be running. Stopping Plex alone is a useful control test only if you also observe the other services.
In a multi-service media stack, Plex can share paths and timing with automation, indexing, and download services.
Capture the process that opens the block device or mount during a wake event, then pause that service for one scheduled cycle. When the wake persists with Plex stopped, keep the diagnosis outside Plex.
Change Sleep Policy Only After the Cause Is Known
Aggressive spin-down can increase start latency and may not be desirable when the server legitimately performs frequent maintenance. The objective is fewer unnecessary wake-ups, not maximum sleeping time at any cost.
utilization, saturation, and error checks separate a busy resource from one that is actually constrained or failing.
After removing an unnecessary wake source, observe a full day of spin-up count, playback start time, and maintenance completion. If useful Plex work now misses its window or playback delays become unacceptable, restore the task or choose a less aggressive power policy.
Support & Tips
More to Read

Live TV Recording Storage Guide for Capacity, Retention, and Cleanup
Measure real recordings, reserve headroom, combine age and capacity limits, and prove the oldest eligible program is removed before storage fills.

Home Media Metadata Recovery Workflow After a Database Restore
Protect the restored state, verify media identity and paths, then repair missing artwork or matches in a pilot library before broad metadata changes.

Jellyfin Client Compatibility Checklist for Audio, Video, and Subtitles
Test representative files one variable at a time and record Direct Play, remux, audio conversion, video transcode, or failure for every client.

