Attempted to migrate pools now failing to start docker

TrueNAS General
1

I attempted to replace my main pool with a new pool. The data all copied over, but now when I attempt to launch the apps I get the error: Application(s) have failed to start: Failed to start docker service

I also have already wiped the original pool thinking the process was successful. I know I messed up there.

Not sure if it matters, but my original install was with hexos. I have since done far more within turenas directly since I wanted to do more complex stuff.

Is there some step I missed, or something is does this mean the whole pool is screwed up. I have 12tb of data on the pool(still readable though windows smb share) but I would like to avoid having to move it if possible.

Update: Google Gemini was able to resolve this with a lot of back and forth. Here is the summary for anyone who stumbles on this.

The issue was a combination of three factors during the pool migration:

ZFS Mountpoint Glitch: After replicating the data to the new pool, the ix-apps dataset had an incorrect mountpoint. I had to manually set the mountpoint to the hidden system path (/.ix-apps) to prevent a “double-mount” (e.g., /mnt/mnt/.ix-apps) that was preventing the Docker engine from initializing.

Stale Binary Datasets: Because the original install was moved from a previous version/configuration, old Kubernetes-era binary datasets (bin, lib) were present in the ix-apps folder. Deleting these allowed Electric Eel’s fresh Docker engine to start with a clean slate.

Read-Only Safety Locks: The replication process left the entire pool and its sub-datasets in a “Read-Only” state. This triggered [Errno 30] errors when the apps tried to start. I had to recursively clear the readonly property and specifically unlock app-specific sub-folders like Plex’s logs and transcode datasets.

Key Commands Used:

  1. Fix Mountpoint: zfs set mountpoint=/.ix-apps <pool>/ix-apps
  2. Unlock Pool: zfs set readonly=off <pool>
  3. Inherit Writable State: zfs inherit -r readonly <pool>

Once the file system was writable and the mountpoints were corrected, I re-selected the pool in the Apps > Settings UI. The system initialized, recognized the existing app_configs, and all my containers successfully deployed.

2

It sounds like you messed up the moving of the applications. Please post detailed listing of you hardware, OS version, pool layouts and what apps you are using.

The data should be okay for the file sharing but guessing the apps need to be set up again. The better the details on your system, the better the advice from our users.

You can expand the Details section of my sig to get an idea of what would be helpful.