Hi All,

I updated to 7.3.4 today and after rebooting, my system was not accessible through the url using my server name (https://scooby.local) or ssh. Some troubleshooting, I found that I could access the web interface via http://tower.local and that a number of setting have been lost. In addition to ssh access being disabled, all of my docker containers are gone, ssh doesn't take my key (after I re-enabled ssh). I was able to start the array, but the loss of the docker containers is painful. Any suggestions on getting things back up? Attached are my diagnostics.

tower-diagnostics-20261030-1145.zip

The diagnostics show that the filesystem on your boot flash was damaged and then repaired. There are multiple FSCK000x.REC files on the flash, which are the pieces left by a filesystem repair. Because of this, several configuration files went back to their defaults: the server name, the SSH settings and key, HTTPS, NTP, the network settings, and the cache pool assignment.

Your containers were not deleted. With the cache pool unassigned, Docker and the VM service created new, empty docker.img and libvirt.img files on disk1. Your original images and appdata should still be on the NVMe device, which Unassigned Devices now shows as unassigned.

If you have a flash backup from before the update, the easiest fix is to restore the config folder from it. That brings back all the settings at once. Otherwise, do the following:

1. Go to Settings > Docker and Settings > VM Manager, and set both to disabled.

2. Stop the array. On the Main page, assign the WD_BLACK SN770 to the cache pool slot and start the array. Do not format anything. If the cache shows as unmountable, stop and post new diagnostics.

3. Check that the original images are back on the cache, in /mnt/cache/system/docker and /mnt/cache/system/libvirt. Then delete the new empty images on disk1: /mnt/disk1/system/docker/docker.img and /mnt/disk1/system/libvirt/libvirt.img.

4. Enable Docker and VM Manager again. Your containers should be back. If any are missing, you can reinstall them with their previous settings from Apps > Previous Apps.

5. Set the server name again in Settings > Identification. Add your SSH key again in Users > root. Check Settings > Management Access (HTTPS), Settings > Date and Time (NTP), and Settings > Network Settings.

When everything works, back up the flash. Then check it for errors on a PC. I also recommend you replace it with a good-quality USB 2.0 flash drive, because Ultra Fit devices have a history of failing. If this happens again on the same drive, it is a strong sign the drive is failing.

If anything does not look right after you assign the cache, post new diagnostics.

That worked! Thanks for the detailed and customized instructions on getting things back up.

Now I need to look into migrating to an internal boot device because this is the 3rd USB drive to fail in as many years. This was after close to decade on my original USB drive finally gave up the ghost.

12 hours ago, BenjaminMetzler said:

Now I need to look into migrating to an internal boot device because this is the 3rd USB drive to fail in as many years. This was after close to decade on my original USB drive finally gave up the ghost.

I'd advise to stay on USB boot.

But get a reliable drive (ideally with a spare) that is designed for always-on operation. It will likely outlast your server.

https://forums.unraid.net/topic/199642-psa-on-industrial-mlc-usb-boot-drives-mlc-nand-end-of-life-and-last-accessible-stock/