`talosctl -n <ip> reset` On RaspberryPi Wipes Firmware

[chr0n1x]

Bug Report

RPi4 Model B that was already running talosOS as a k8s-worker node fails to reboot/startup after a talosctl reset

Description

I have my baremetal cluster running on a few RPi4s, model B.

While experimenting with things, I decided to reset one of the nodes before applying a new machine configuration. The command run was:

talosctl -n <node IP> reset

Logs

The command exit 0s on my main machine. The talos rpi worker however just seemingly goes into the rpi poweroff state. After unplugging the power and plugging it back in with a display, The RPi4 ModelB boot screen is present (instead of the uboot one) and there's a Firmware not found message.

Environment

• Talos version: 1.7.5
• Kubernetes version: 1.30.1
• Platform: Raspberry Pi 4 model B. Host running talosctl is wsl2 ubuntu on win11 (if that matters)

[frezbo]

Yes a normal reset would wipe the whole disk, if you just want to reset back to maintenance use --system-labels-to-wipe=STATE --system-labels-to-wipe=EPHEMERAL