Happy new year!

Wishing everyone a New Year filled with good health!

I am using latest Unraid. I created a single btrfs cache pool with two disks. I set spin down in 15 minutes.

I don't use them anywhere in unraid. Dockers/VMs...

I want only to use them to rsync a share /mnt/Photos once per week.

For some reason they spin up as it appears in specific intervals.

I run File monitor and it didn't catch anything as expected so there must be either unraid or a pluging.

I have attached diagnostics.

I will uninstall some plugins as a first step. (I was trying to make use of the igpu of the intel ultra 2 but I think I need to wait for the v7.3 with new kernel).

Any suggestions what/where to look?

tower-diagnostics-20260101-0945.zip

I still cant figure what causes the disks to spin up.

I uninstalled some plugins but it didn't fix it.

Any idea what to look for?

tower-diagnostics-20260105-1007.zip

P----s                            shareUseCache="no"      # Share exists on disk1, disk2, disk3, photoscopy

Assuming that pool is named "photoscopy" it looks like you made your copy use the same share name, so anything that reads from the (I assume) "Photos" share will have to list its contents too.

Should use a different share name and set it to only live on that pool.

I have not created any share!

I only created a new cache pool using two drives as single and I used their mount point /mnt/photoscopy to rsync my share /mnt/Photos

It drives my crazy! I uninstalled most of my plugins, checked all mount point in running dockers (VM are shut down) and checked all settings in Unraid for any scheduled job but I didn't find anything!

There are known issues with btrfs pools spinning up, there is one with zfs, suggesting something must be using it, recomend booting in safe mode to rule out any plugins.

33 minutes ago, karateo said:

I only created a new cache pool using two drives as single and I used their mount point /mnt/photoscopy to rsync my share /mnt/Photos

That's the whole point, this will make the copy appear as part of the same share (a share is just "any top level folder on any array or pool drive with that name"), so anything accessing the "Photos" share will wake up that pool to list its content as a result.

Make a new share on that drive and copy the contents of the Photos share to that new share on that pool instead.

27 minutes ago, Kilrah said:

That's the whole point, this will make the copy appear as part of the same share (a share is just "any top level folder on any array or pool drive with that name"), so anything accessing the "Photos" share will wake up that pool to list its content as a result.

Make a new share on that drive and copy the contents of the Photos share to that new share on that pool instead.

1) Creating a new cache pool using two drives as single with a different name (/mnt/photoscopy) than any other share is not a problem right?

2) How can this command appear as a copy? It's not linked in any way with the old share (/mnt/user/Photos).

rsync -avh --info=progress2 --stats /mnt/user/Photos/ /mnt/photoscopy/Photos/

The new mount point should be "incognito" !!!

I will try to create a new share using Unraid GUI and select only the new cache pool and I will run

mv /mnt/photoscopy /mnt/new-share-name

right?

44 minutes ago, JorgeB said:

There are known issues with btrfs pools spinning up, there is one with zfs, suggesting something must be using it, recomend booting in safe mode to rule out any plugins.

I will try what Kilrah suggests and if I can't solve it, I will run in safe mode.

I need Homeassistant to run my house so this is the worst case scenario!!!

6 minutes ago, karateo said:

/mnt/photoscopy/Photos/

All top level folders will be treated as part of a User Share of that name regardless of whether the Share settings actually mention that drive.

In addition the command you give mixes a User Share and a physical device in the same command. This is dangerous and can lead to data loss if you are not very careful.

Just now, itimpi said:

All top level folders will be treated as part of a User Share of that name regardless of whether the Share settings actually mention that drive.

In addition the command you give mixes a User Share and a physical device in the same command. This is dangerous and can lead to data loss if you are not very careful.

I created a new share with only new cache pool named Photosimage and run mv /mnt/photoscopy/Photos /mnt/photoscopy/Photosimage/Photos

I don't know why the cache only Photosimage has a green cycle thinking thats in the array!

48 minutes ago, karateo said:

I don't know why the cache only Photosimage has a green cycle thinking thats in the array

The green circle does not mean that the Unraid thinks the file is on the array. Instead it thinks that it is on redundant storage and is protected.

At the moment Unraid is not taking into account the profile used on the pool and automatically assumes that a multi-drive pool is protected. I believe that a future Unraid release is expected to take the profile into account.

17 hours ago, Kilrah said:

That's the whole point, this will make the copy appear as part of the same share (a share is just "any top level folder on any array or pool drive with that name"), so anything accessing the "Photos" share will wake up that pool to list its content as a result.

Make a new share on that drive and copy the contents of the Photos share to that new share on that pool instead.

I think that solved it!

I didn't know that

/mnt/user/Photos

appears the same folder as

/mnt/some-cache-drive/Photos

Thank you for your help. It drove me crazy as there was nothing accessing the cache pool directly!

No spin ups since yesterday afternoon that I did what you suggested.

Thank you all!

Edited January 9Jan 9 by karateo