I have 4 Raspberry Pi running DietPi. Today I ran the command
sudo apt update && sudo apt full-upgrade -y
as I usually do to update. After the update I proceeded to restart the RPis just to not been able to SSH into any oft them. Any of the services running stoped working. I had 2 RPis running Pi-hole, 1 RPi running WireGuard and 1 running Homebridge. I then went and flash all of them with a DietPi in order to start from scratch. I can SSH into them. Run the initially setup. After it’s done and I reboot (I’m guessing it updated everything during the install), I can’t SSH into it again. I tried this with two different RPis and same result. At the time this happened I was (am) running Dietpi 8.0.2 64-bit.
Any suggestions as to what could be breaking? It seems to happened in the latest APT update/upgrade.
I am having the same issue. My pihole box is now non-bootable. Luckily I do have a backup of the pihole configuration so I can restore it.
One thing to note though is that this also affects new installations since Dietpi will automatically run apt upgrade during installation. Anyone trying to install it now is getting a dead system (unless they disconnect their pi from the internet).
It looks like apt update & upgrade is done automatically on my Pi, luckily it did not grab that faulty kernel (probably due to the timezone I was in). Thinking back, I think it was PiVPN Wireguard that configured the system to check for updates every day, is there a way to undo this change? Or at least change it to weekly?
t looks like apt update & upgrade is done automatically on my Pi, luckily it did not grab that faulty kernel (probably due to the timezone I was in).
yes luckey RPi guys removed the kernel package quickly after our reports.
I think it was PiVPN Wireguard that configured the system to check for updates every day
yes a bad behaviour of PiVPN. But if I’m not mistaken we suppress this in meantime.
In your case there should be unattended-upgrades tool running. I will link the Debian Wiki. There you have a section about timers who should manage the execution.
in meantime issue has been identified by RPi developer. Unfortunately it was a rar combination of a setting we do on DietPi and how RPi kernel was handling a specific situation. A kernel fix is on the way and I already tested it on my demo system
root@DietPi4:~# uname -a
Linux DietPi4 5.10.92-v8+ #1514 SMP PREEMPT Mon Jan 17 17:39:38 GMT 2022 aarch64 GNU/Linux
root@DietPi4:~#