Hi all,

I'm running nextcloud-aio-mastercontainer on Unraid and trying to optimize for NVMe cache usage while keeping bulk data on HDDs.

Setup:

- Nextcloud AIO Docker

- Data directory: /mnt/user/nextcloud

- Share configured with cache + array

Problem:

Even when data is physically located on the NVMe cache pool, my HDDs still spin up frequently during normal Nextcloud usage.

This happens during:

- Web UI browsing

- File listing

- Preview access

- general metadata operations

What I observed:

Even without accessing HDD-resident files directly, disks still wake up.

My assumption is that Unraid's FUSE layer /mnt/user) is resolving file metadata across all disks, not strictly using cache-only paths.

- HDDs do not stay spun down

- NVMe cache benefits are partially negated

- noticeable increase in disk activity and power usage

Is there any recommended way in Unraid to:

- fully isolate Nextcloud data to cache for metadata + IO

- avoid /mnt/user FUSE overhead for this workload

- or split hot metadata (previews/appdata) and cold user data safely?

I'm especially interested in setups that still work cleanly with nextcloud-aio-mastercontainer

Thanks in advance!

You are likely to get more informed feedback if you attach your system's diagnostics zip file to your next post in this thread.

It is always a good idea to do this to allow us to see the current state of your system and so we can see logs and configuration information.