[6.12.6 => 6.12.8] Some settings reset after update & Docker disabled

Quote

Also an unscheduled parity check started by itself.

 

This suggests there was an unclean shutdown, and the docker settings have be known to sometimes revert to defaults after an unclean shutdown.

Diagnostics are required for bug reports. 

 

 

1 minute ago, JorgeB said:

 

This suggests there was an unclean shutdown

 

The power never stopped, I never unplugged the computer and I didn't use the physical buttons on the machine, that's why I thought I'd post here.

 

I did two reboots, so not sure when Docker broke:

- one because I was unable to provision a SSL certificate, so I did a manual restart from the Main page.

- one after upgrading the version, again from the main page.

 

Nothing meaningful (for me) in the logs, these are the only errors and warnings from yesterday evening.

WARN: failed to prefill DIMM database from DMI data

WARN: ACPI: Early table checksum verification disabled

ERR: unraid sshd[1112]: Received signal 15; terminating.

ERR: unraid sshd[1112]: Server listening on 192.168.1.195 port 22.

3 minutes ago, trurl said:

Diagnostics are required for bug reports. 

 

Attached.

 

 

unraid-diagnostics-20240220-1240.zip

Edited February 20 by Nedevski

1 hour ago, Nedevski said:

The power never stopped, I never unplugged the computer and I didn't use the physical buttons on the machine, that's why I thought I'd post here.

There was still an unclean shutdown:

 

Feb 19 23:12:07 unraid emhttpd: unclean shutdown detected

 

The fact that settings are reset, and also unclean shutdown, makes me suspect flash drive corruption, since all of your settings are on flash, as well as the status that says the array was stopped.

 

If that status can't be written for some reason, unclean shutdown is the result.

 

And if any settings can't be read from flash, they go back to defaults.

 

What do you get from command line this?

ls -lah /boot

 

2 minutes ago, trurl said:

What do you get from command line this?

ls -lah /boot

Actually, I can see this from diagnostics.

 

You have a lot of FSCK files on flash, which are the result of repairing corruption on flash.

 

Do you have a current backup of your flash drive?

 

You should probably reformat flash, reinstall, and copy your config from your flash backup.

On 2/20/2024 at 4:27 PM, trurl said:

You should probably reformat flash, reinstall, and copy your config from your flash backup.

 

I'm a bit scared of doing that, I've just setup everything, including Docker containers  But if I have to...

 

On another note, it's a new genuine 32GB Phillips USB2.0 flash drive used in a USB 2.0 port, so it *should* be fine?

 

I have cloud backup enabled, so I should be fine in the event of flash drive failure.

I will also look into reflashing and might try it out, thank you!