NS8: after adding disk no full boot

Support
,
1

NethServer Version: 8
Module: none
Hi all,

I am setting up a NS8 virtual machine using the rocky linux qcow image.
All goes well, untill I add an extra disk /dev/sdb, do

parted /dev/sdb -- mklabel gpt
parted -a opt /dev/sdb -- mkpart primary xfs 0% 100% 
mkfs.xfs /dev/sdb1
mkdir /mnt/temp_disk
mount /dev/sdb1 /mnt/temp_disk

and use the script from this page to migrate /home to the new disk.

From then on, the VM won’t boot, stops on " Probing EDD"
I’ve read quite some similar errors, but I cannot fix it.
What I’ve tried:

  • checked CPU type (yes, it is x86-64-v2-AES)
  • after adding the line in /etc/fstab as suggested by the script, I ran systemctl daemon-reload (as is said in the fstab file, not as script output)
  • I update grub-config (rub2-mkconfig -o /boot/grub2/grub.cfg )
  • using blkid, I see no change in UUID’s of disks
  • in grub, change console=ttyS0 into tty0 (using e while booting), gives me the option to continu booting.
    When I then log in, and comment out the line added to fstab, I can reboot succesfull.
    So I guess, it boils down to: what is wrong with this fstab line:
    UUID=f8764d23-e317-4ffc-98af-dbaf75a8504e /home xfs rw,errors=remount-ro 0 1
    (NB: this is the line suggested by the script).

Any help appreciated. If more info is needed pleaselet me know.

root@node scripts]# blkid 
/dev/sdb1: UUID="f8764d23-e317-4ffc-98af-dbaf75a8504e" TYPE="xfs" PARTLABEL="primary" PARTUUID="1ea35579-6aa9-44d3-beaf-51d787dfd314"
/dev/sda2: LABEL="boot" UUID="25588c14-4705-4f05-ab0b-e6cde771de92" TYPE="xfs" PARTLABEL="primary" PARTUUID="47f7a431-5f58-4d24-a1d7-921c83d6f9de"
/dev/sda5: UUID="7a97e869-e298-4741-8c4b-8b1b714435e5" TYPE="xfs" PARTLABEL="primary" PARTUUID="cc974f06-f3a8-433f-a5bf-dbed9f9d2567"
/dev/sda1: SEC_TYPE="msdos" UUID="89CA-9A88" TYPE="vfat" PARTLABEL="EFI System Partition" PARTUUID="39d4ce0b-8cc4-4b71-bfc7-df0c0a16fa48"
/dev/sda4: PARTLABEL="primary" PARTUUID="4d2d5504-6634-4335-bcc9-9a2bc0960504"
/dev/sda3: PARTLABEL="primary" PARTUUID="07231053-988f-4679-84c8-03ddc8791c1b"

Edit: took out the “EDD” part from the title, as not relevant, just the last line showing.

2

OK, think I found the problem.
The output if the migrate-home script reads:

"add to /etc/fstab:
UUID=f8764d23-e317-4ffc-98af-dbaf75a8504e /home xfs rw,errors=remount-ro 0 1 "

When booting, hit e to edit boot options, change console=ttyS0 into tty0.
Then I could read boot logs, and log in in " emergency mode" .
There I saw:
kernel: xfs Unknown parameter 'errors' .
After removing that parameter from the fstab line, all goes OK!
In short: param “errors” is not supported for XFS (in current version?)

Then, a follow-up question:
I now boot with only “rw” as param.
However, / (also XFS) is mounted with “defaults” . Should I mount /home also with 'defaults" ? Or is rw sufficient?
Also, should the script be adjusted to correct this?

3

Both should work the same way: