MemTest on boot!!??

Just updated to OMV7, couldn't work out why the reboot was taking so long, connected a monitor and it's running a memtest??? Tried a keyboard and hit esc but the machine just restarts and goes back to the memtest.

The initial grub screen shows no reference to the pve kernel just the memtest option, fortunately I have a boot backup of OMV6 I can deploy to another flash drive, but running another omv-release-upgrade from that does not encourage me.

Quote from ryecoaaron

Do you have the upgrade output

No unfortunately, run from a windows command line, but didn't see any errors on screen.

Quote from ryecoaaron

I would just boot the debian netinstall iso and choose rescue or systemrescue and choose to boot the disk from its grub install. Then you can fix grub from the install.

OK will have a look at that, probably be Friday before I get the time.

Nope tried the rescue install option doesn't work and I know my boot device is /dev/sdg as it has 3 partitions on it and /sdg always appears last in the drive list on my OMV. Tried the rescue three times and it still goes to memtest

+ I can't get to the usb backup file because it's on one of the drives in the server and it's formatted to ext4

So two options,

1) I deploy a linux distro on an old laptop, remove the drive from my server which means removing it from the straps holding it to the underside of my desk, remove the drive and hope I can get the usb backup completed on Sunday copied to another flash drive, so I can flash it to another usb flash drive

2) Total clean install, which would mean reconfiguring everything, docker, shares, rsync etc and hope my zfs pool imports OK

Not happy!! I don't really have the time nor the patience for this at the moment

I still can't get this to work, I have disconnected 5 of the 6 drives, the sixth drive is connected internally, 4 are caddy drives and 1 is esata to sata.

I can boot using the net install, select advanced, then rescue mode, if I then execute a shell and run blkid I can see the internal drive, so note that /dev/sda1, one usb is clearly the net installer, so the other is omv's boot drive /dev/sdb5 is the swap and I assume /dev/sdb1 is the boot partition.

Exit shell, go back to the option to select device to use as root file system, select /dev/sdb1, back again, install grub boot loader, all this appears to work but I'm obviously doing something wrong as the boot loader displays, but no kernel just memtest

So how do I 'inject'/add a kernel to grub

Quote from ryecoaaron

And just to make you not hate OMV or Linux

It's not that, I'm approaching a first anniversary of an event from last year so I am somewhat all over the place, my computers keep me occupied

I spent a day last week trying to find out why my Rasp Pi LibreElec wasn't running, after trying different things turned out to be the power supply, fortunately as it's a 3B I did find something that matched the required output, then a reinstall and a reinstall of the Emby addon......success

Quote from ryecoaaron

Did you happen to run update-grub

From the net install? not sure if there's an option

Quote from ryecoaaron

Did you see kernels in /boot

OK, what I have done is added my actual OMV boot usb to a VB running OMV7, two images one shows /boot and the other the pve folder, no kernels

Quote from ryecoaaron

You might have to temporarily update /etc/resolv.conf to access the internet

it's blank, and it doesn't allow me to write to the file to add nameserver 1.1.1.1 -> error no such file or directory

Been able to get the system back up using a previous omv6 usb have copied the backup file to my workstation where balena etcher is writing the image to another usb, I'll see what happens with that

The restored backup is working but my zfs pool has gone, it won't import, it show the folder in the root directory, but the filesystem/folders show empty in file browser

ok further investigation, the four drives don't show in storage -> disks......this is so screwed

Quote from ryecoaaron

Are they still disconnected

Yes cos I'm an idiot and didn't check that all hotswap drives were correctly seated and their trays in the upright position

So back square one, fortunately I did not pass go, it's getting late here across the pond, proposal...........

If I do this again just in case it goes belly up is there anyway I can record the output, or just copy and paste the output into a txt file

This time it worked, the output from each update-grub showed a kernel present -> new grub

update-grub
Generating grub configuration file ...
Found linux image: /boot/vmlinuz-6.2.16-20-pve
Found initrd image: /boot/initrd.img-6.2.16-20-pve
Found memtest86+x64 image: /boot/memtest86+x64.bin

The only error I saw was this:

Updating workbench configuration files ...
Restarting engine daemon ...
Errors were encountered while processing:
 grub-pc
E: Sub-process /usr/bin/dpkg returned an error code (1)

I also had to run the the fix6to7upgrade script as I was getting a 502 bad gateway trying to login to the GUI

So the only difference was deploying the usb-backup image to one of my OMV system backup flash drives

Edit: My proxmox kernel is 6.2 should I update it to 6.8

Quote from ryecoaaron

Did you get grub-pc fixed

No, seems to be a number of possible fixes with the main one being dpkg --configure -a

EDIT: that is the command to use, just had the error appear installing updates with the solution as above