Im really in a bind here. For some reason overnight i started getting btrfs rebuilding errors on my 2 drive cache array (ssd). I decided to move everything back to the  parity protected array (spinning disk)

 

I probably messed something up here.  would really appreciate assistance. These are the steps i took to move everything off of cache to parity protected array

1. turn off docker / vm in settings.

2. move the folders that were on /mnt/cache/ to /mnt/disk1/  (appdata, domains, system) via MC

3. set shares to no for using cache disk

4. turn off array

5. remove the cache drives from the array

6.  start array. 

 

if i go to /mnt/user i can see all the folders and can access the files

 

Thank you for your time and help!

merlin-diagnostics-20170524-0816.zip

What are your settings for libvirt and docker image storage locations?

I am not quite sure what the problem is?    Once you moved all the folders from the cache to disk1 and removed the physical cache disks then for them to be showing up under 'mnt/user is normal behaviour!     The /mnt/user folders are simply a merged view of all the top-level folders on the array disks.

7 hours ago, jonathanm said:

What are your settings for libvirt and docker image storage locations?

 

Docker storage location: /mnt/user/system/docker.img
Default appdata storage location: /mnt/user/appdata/

Libvirt storage location: /mnt/user/system/libvert/libvert.img
Default VM storage path: /mnt/user/domains/

What are the chances that i have corrupt image files? and what would it take to rebuild the image file? 

i have backups of all my vm's and docker appdata. 

 

Are the errors the image files or the cache disk. The docker.img file is a virtual BTRFS disk file so it can give issues. If it's the docker.img file then you basically stop Docker, delete the image then start Docker again. On the Docker page click the add Docker button then pick the my-xxxx templates to re-install them. Since the application data is stored independently from the docker.img file they come back working with your current setup.

The errors were from the cache disks because it listed sdg an sdh as the source of the rebuilding errors. 

 

Spent the weekend rebuilding the dockers and vm configs, i couldnt find a way to get the existing img files to work properly. Luckily i didnt have to tweak the settings too much to get everything working.  

I precleared / erased (3 cycles) the cache drives and put them back into production, since Friday there hasnt been a single BTRFS rebuilding error.  Starting off slow, ive only moved over appdata to cache to see how it reacts. 

Ive also decided to put the system folder into my backup routine. I didnt do that before, and that was an oversight on my part. Never realized how important those files were.