Recover corrupted HA OS

I have a R Pi4 in a Argon One case with an nvme. The fan in the case failed so I took the case apart. When I put it back together, the system appears as if the OS/ drive is corrupted ( I did a full system shutdown prior to pulling power, so not sure what happened though I had issues the last time I took the case apart).

I connected a monitor and keyboard to the pi it’s showing error loading journal and cannot connect to the Docker daemon... errors. then gives a [WARN] Home Assistant CLI not starting! Jumping into emergency console...

As the emergency console is /bin/ash, I can’t seem to really do anything in the console.

I tried loading an SD card with rasp pi OS then running checks on the nvme in a external case, but it tells me I don’t have permissions to the partitions.

I’m not sure how to recover at this point and unfortunately I’ve done a LOT of work recently since I moved a backup off-disk ( I know, I feel crappy enough right now, so please spare the lectures on this).

I’m definitely not good with Linux so maybe I’ve missed something? Does anyone have a suggestion to restore the system or at least gain access to the files?

Thank You in advance.

Booting on an external SD card with RPI OS seems to be the way to go.

Make sure your nvme partitions are not mounted when you run a disk check, and you are running the relevant commands as root (that is, with sudo).
If you’re lucky you can just mount the partitions, then copy the files as root to extract a backup.
However you may need new hardware eventually if the source of your issues is that you damaged it while taking it apart… But maybe not, who knows!

It looks like the kind of thing that chatgpt would be good at providing detailed guidance on.

I’ve done a LOT of work recently since I moved a backup off-disk

I’m not sure what you mean by this.

I’m running into several issues trying to recover the drive. it tells me that it can’t read the superblock but I’ve tried using the backups and it gives super block needs_recovery flag is clear, but journal has data then usually gives an input output error. I’ve also tried running different fsck but get same error and something like unexpected inconsistency; run fsck manually.

Is there a way to pause to boot screen? I’m trying to read what is failing on boot but it goes by so fast I can’t read it.

Once you’re booted you can use dmesg to get the same log that’s normally printed at boot.

For the rest again, I think chatgpt will be very good at helping you with that :blush:
Good luck!

I’m in a boot loop and can’t fully boot. I think it’s getting to the docker app engine and restarting.

ChatGPT hasn’t been all that helpful through this.

Yeah I mean it should help you with the data recovery commands and errors, not with getting it to boot.

If you have a Windows PC, you might be able to copy the config directory and the backups of the disk

I use it for that, but I don’t always trust what it says. I have gotten wrong or outdated info more than once; that’s why I ask here as well with stuff like this.

Also, that’s not meant to sound snarky because i do appreciate the help. I just wanted to state that as sometimes it’s hard to tell tone in text. :slightly_smiling_face:

I’ll look into this, thanks. I’ve used EXT2 Volume Manager in the past, but since switching to windows 11, it doesn’t seem to want to work.