[solved] Cannot change cache pool of my existing user shares

[trallnag]

Hello, here is the summary: I have two pools. I want to change the cache pool of a user share to use a different pool. I can select the pool I want, but the share continues to use the old pool.

 

Version 6.9.2 2021-04-07

 

I have two pools. One is a SSD, the other one is a NVME.

 

 

I have the default user shares.

 

 

By default they are all using the "General" cache pool. I want to switch all of them except "isos" to the "Speed" pool. When I edit the shares I can select the cache pool I want. But the change is not performed. The selected cache is shown correctly, but when you look at the computed size you can see the shares are still using the 500 GB SSD and not the 1 TB NVME.

 

How can I force the update? Maybe restart the array? What I did for "domains" is to just delete the share and then recreate it.

 

Thanks in advance

unraid-diagnostics-20210506-2146.zip

Edited May 6, 2021 by trallnag

[JonathanM]

I believe the issue is that as of 6.9.2 and earlier the mover only operates between a pool and the main array. So, to move from one pool to another, you will need to set cache:yes on the share, make sure there are no open files, and run the mover. Then once all the files are on the main array, you can set cache:prefer to the pool you want to use and once again run the mover.

 

Alternatively you could move the files manually using mc or something like that.

[trallnag]

I set all the user shares to "yes" and started the mover. Now "appdata" disappeared... Not just from the GUI but also from /mnt/user. It is / was completely empty.

 

I then recreated appdata pointing to the NVME pool and "only" mode, but "Compute" shows the whole disk array as size. Weird.

 

 

May 6 22:28:24 unraid emhttpd: Starting services...
May 6 22:28:44 unraid emhttpd: shcmd (3366): /usr/local/sbin/mover &> /dev/null &
May 6 22:32:54 unraid emhttpd: shcmd (3375): mkdir '/mnt/user/appdata'
May 6 22:32:54 unraid emhttpd: shcmd (3376): chmod 0777 '/mnt/user/appdata'
May 6 22:32:54 unraid emhttpd: shcmd (3377): chown 'nobody':'users' '/mnt/user/appdata'
May 6 22:32:54 unraid emhttpd: Starting services...
May 6 22:33:04 unraid emhttpd: cmd: /usr/local/emhttp/plugins/dynamix/scripts/share_size appdata ssz1 general,speed
May 6 22:33:41 unraid emhttpd: cmd: /usr/local/emhttp/plugins/dynamix/scripts/share_size appdata ssz1 general,speed

 

[trallnag]

1. I activated logging on the mover.
2. Put a dummy file into every share, since some were empty.
3. Set the shares I want to update to "Yes : General".
4. Ran the mover
5. Set share "appdata" to "Prefer: Speed"
6. Run mover
7. Set share "appdata" to "Only : Speed"

Now it looks correct! Thanks!

 

Regarding "system" it seems like it cannot move a few files and so the pool is not updated. 

 

May 6 22:46:49 unraid root: mover: started
May 6 22:46:49 unraid move: move: file /mnt/general/appdata/lol.txt
May 6 22:46:49 unraid move: move: skip /mnt/general/system/docker/docker.img
May 6 22:46:49 unraid move: move: skip /mnt/general/system/libvirt/libvirt.img
May 6 22:46:49 unraid move: move: file /mnt/general/system/lol.txt
May 6 22:46:49 unraid root: mover: finished

 

[itimpi]

16 minutes ago, trallnag said:

I then recreated appdata pointing to the NVME pool and "only" mode, but "Compute" shows the whole disk array as size. Weird.

us9ng Only will not get the files that are on the array moved to the pool as mover ignores any shares which are set to Only (or No).  You need Prefer go make this happen.

[itimpi]

1 minute ago, trallnag said:

Regarding "system" it seems like it cannot move a few files and so the pool is not updated. 

Did you make sure the Docker and VM services were disabled while you wanted the System share moved?   The services will hold files open thus stopping mover from taking action on them.

[trallnag]

Ah yes, that was missing. Now it also worked with "system" share.

 

Edited May 6, 2021 by trallnag