Latest update killed HA?

I cannot recall the version anymore, but I’ll walk you through the steps I took if it helps much. I got notifications that the version I was using was insecure and that it was recommended to update. Went to supervisor and there were 2 updates available. Got them updated, but then HA just dropped off the network entirely. I checked on my router and it did not show up (Wired connection). Plugged the HDMI cable into my monitor and I see 5 EXT4-fs errors;

3 of them on inode #33, #34 and #31 said:

htree_dirblock_to_tree:997: inode [number]: comm systemd-tmpfile: Directory block failed checksum

2 of the on inode #31 and #34 said:

__ext4_find_entry:1541: inode [number]: comm systemd-tmpfile: checksumming directory block 0

I’m wondering if my SD card was corrupted somehow?

What would be the best way to restore what I had previously. I had literally just made a snapshot before updating, but realised after the fact that I forgot to download it…go figure… :man_facepalming:

Is a fresh start my only option now?

Well, booting from the SD is probably out of the question if the checksums are toast, but you can try to connect the card to a computer and see if you can salvage your snapshot that way, otherwise you may be in fresh start territory.

The system is complaining about the temp file, so perhaps that is the only smoked directory - meaning your snapshot may still be accessible if you throw that card into a card reader and access it on your computer.

Silly question here, but I use a mac and the card becomes unreadable on my machine (even when freshly flashed). Would I need to set up a linux system with virtual box or something similar in order to access these files? Even just my config file would be nice to have.

Take it you are using a Pi ? Have you got login access ? If so login and enter:-:
sudo fsck -A -C -l

Then reboot

EDIT: Just remembered, this wont help because the mmcblk0o2 block device for / will be locked (mounted). So, add the following to /boot/cmdline.txt :

fsck.mode=force

This has to be on the same existing line as the boot command. Likely to have
fsck.repair=yes already present. Leave it and/or if it not there then add it with the above.also - but separate the two with a space.

I’ve set up HA on another SD card in the mean time starting from scratch, but I’ll give this a go! Thanks!

Rather extreme, but your call. For info, the Pi filesystem is such (well by default ‘ext4’) that multiple copies of I/O file level attributes key to integrity, are stored and managed. A very basic summary but what’s important is that the system can recover from bad SD cards or power cuts (including accidental switch off). Running that command on boot will attempt to resolve the system. If you’re unlucky you might loose a few files out of thousands that stops you working - although that has rarely been the case for me.

(JFI - it took Microsoft around 3 decades to catch on to this and build a resilient OS aka Windows 10) :grinning: