System: LittleBoy, just upgraded from 7.2.x to 7.3.0 today.

Pre-existing setup: LUKS-encrypted parity array (disk1–disk11), plus btrfs pools (cache, blitz, vm, misc, black), a ZFS RAIDZ1 pool (tempspin), and a half-created encrypted ZFS RAIDZ1 pool (oldvault) that I was setting up just before the upgrade.

After upgrade, I tried to use a 60GB SATA SSD (sdad in current device naming) as an internal boot device via the new 7.3 wizard. The wizard tried to format the existing tiny sdad4 partition (1.3MB) and failed with "ERROR: no btrfs on /dev/sdad4" and "boot: mount error: wrong or no file system."

I wiped the partition table with sgdisk -Z /dev/sdad to give Unraid a blank disk. On the retry, emhttpd segfaulted (segfault at 518, in emhttpd binary). Restarting emhttpd via /usr/local/sbin/emhttp brings it back briefly but it crashes again.

Currently all disks show in Unassigned Devices. /boot/config/super.dat dated 11:06 today (pre-upgrade) still exists, no .bak. Data on disks is intact at the LUKS layer.

Diagnostics attached.

littleboy-diagnostics-20260524-1507.zip

18 hours ago, lakr said:

Restarting emhttpd via /usr/local/sbin/emhttp brings it back briefly but it crashes again.

This is not supported. Reboot the server, boot from the flash drive, and try recreating the boot pool