[SOLVED] Shrunk array, lost all dockers and VM?

[Chuck Claunch]

Hi, I just followed the directions (for unRAID v6.2 and later) on this page to remove a single 4TB drive from an array which originally had 3 4TB drives (2 + 1 parity).  Everything seems to be going as expected, and the parity is currently rebuilding.  Is it normal to expect that all my dockers and VM would be lost in this transition?  Or will they all re-appear once the parity sync is finished?  That's a huge bummer if I've lost it and the instructions had no real warning about loss of data.  I did back up a lot of important data but did not back up my docker config as I thought that lived on the flash drive.  The only other thing I notice is I removed a completely unused cache device (one remaining) and the remaining one says `Unmountable: No pool uuid`.  Any thoughts would be appreciated.

[trurl]

If the disk you removed had no data on it then the likely cause of your docker problems is that unmountable cache. The only part of dockers that are on flash are the templates, which contain the entries for the Add/Edit Container page such as your mappings, etc. The docker.img, which contains the executable code of all of your containers, is best kept on cache, along with the appdata.

 

If possible before rebooting and preferably with the array started
Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

[Chuck Claunch]

Here's the file.  Note the cache device I removed is the PNY_CS1311_120GB, the one I kept (which I believe may have my docker stuff on it) is the SAMSUNG_SSD_830.  Also note, the Unassiend device ST10000VX0004 a drive set up only for a windows VM (boot + data), which I also seem to have lost, though that drive mounted and is readable.

unraid-diagnostics-20201212-1828.zip

[trurl]

Your appdata, domains, and system shares were cache-prefer so they would have been on cache but have now been created from scratch on disk1.

 

When you first booted over a week ago you had parity, disk1, disk2, and 2 cache disks. Then, you did New Config without disk2 to shrink the array, but it looks like you also only left one disk assigned as cache.

 

Did disk2 have any data on it?

 

It is unable to mount cache because of this:

Dec 12 16:21:52 unraid emhttpd: /mnt/cache mount error: No pool uuid

I'm not entirely sure how to proceed from there. You may have to wait till morning and see if @JorgeB can help.

 

In the meantime you should disable both Docker and VM Manager in Settings and get rid of appdata, domains and system shares on disk1 since you need to have it use those on cache when you get cache fixed.

[Chuck Claunch]

I left the one disk assigned as cache because it appeared as if the other one never got used at all (no data on it).  I suppose I could re-assign it just to see if everything comes back?  I can try that once parity finishes syncing I suppose.

[JorgeB]

Try this:

 

Stop the array, unassign all cache devices, start array to make Unraid "forget" current cache config, stop array, reassign both cache devices (there can't be an "All existing data on this device will be OVERWRITTEN when array is Started" warning for any cache device), start array and post new diags.

[Chuck Claunch]

Ok as I was going to unassign the cache devices as you said, I noticed I'd left the slot number at 2.  When I changed it properly to 1 and started the array it loaded the cache drive just fine, then after re-enabling docker and VMs, and a reboot, everything is back up.  Thanks a lot for the help!