sithuk / ubuntu-server-zfsbootmenu Goto Github PK

I have 2 SSDs in my laptop that I stripe across (not enough storage to mirror). It would be good to have an option for striping across devices, apart from mirroring and raidz.

Could someone write down the key bullet points for a safe 22.04 -> 22.10 upgrade?

0. snapshot everything
1. cloning rpool/ROOT/ubuntu-whatever

But what then? Is there anything needs to be done in grub/dracut or must happen in some specific order, or literally just: reboot & automatically I can chose to boot from the new clone, can perform a normal do-release-upgrade and, if system is working as expected: be happy. If not, boot next time again from the "old" root pool and everything will be fine as before the do-release-upgrade attempt ?!

Don't want to mess the weekend and maybe, some kind soul exactly know what todo and could tell?

[edit] coming from a system installed with this script (22.04)

This package doesn't seem to be needed (isn't included in deskop installs normally) and makes problems

Hi,

No issues, just wanted to give you kudos for the work you've done. !!

I have needed to modify your script substantially because my requirements are simple, but let me know if you want feedback from this noob.

Sam

Post reboot, and post installation, I'm trying to set up remote access to zbm,and it's failing.

Here's the output when I run ubuntu_server_encrypted_root_zfs.sh remoteaccess:

[root:~] # ./ubuntu_server_encrypted_root_zfs.sh remoteaccess
Boot environment check passed. Found EFI boot environment.
Mon Feb 20 09:17:20 AM EST 2023
Running remote access to ZFSBootMenu install. Press Enter to Continue or CTRL+C to abort.

***WARNING*** This script could wipe out all your data, or worse! I am not responsible for your decisions. Press Enter to Continue or CTRL+C to abort.


WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

Reading package lists...
Building dependency tree...
Reading state information...
dracut-network is already the newest version (056-3).
dropbear is already the newest version (2022.82-4).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
fatal: destination path 'dracut-crypt-ssh' already exists and is not an empty directory.
mkdir: cannot create directory ‘/usr/lib/dracut/modules.d/60crypt-ssh’: File exists
cp: -r not specified; omitting directory '/tmp/dracut-crypt-ssh/modules/60crypt-ssh/helper'
Generating public/private rsa key pair.
/etc/dropbear/ssh_host_rsa_key already exists.
Overwrite (y/n)? y
Your identification has been saved in /etc/dropbear/ssh_host_rsa_key
Your public key has been saved in /etc/dropbear/ssh_host_rsa_key.pub
The key fingerprint is:
SHA256:lgVpBeE9O829TQRoRXdK/eWVRhDAbMJbbPl/ZMUpNOo root@mishac
The key's randomart image is:
+---[RSA 3072]----+
|        ===.+OO+=|
|       .o= O+o.BB|
|       .. Xo. +.*|
|         +.= o .=|
|        S oEo oo.|
|       .   .   =.|
|              . o|
|                 |
|                 |
+----[SHA256]-----+
Generating public/private ecdsa key pair.
/etc/dropbear/ssh_host_ecdsa_key already exists.
Overwrite (y/n)? y
Your identification has been saved in /etc/dropbear/ssh_host_ecdsa_key
Your public key has been saved in /etc/dropbear/ssh_host_ecdsa_key.pub
The key fingerprint is:
SHA256:6CjrLvtymb6YAF3sPNKT82/0lvIMD9Yx4VvxCt8mgi8 root@mishac
The key's randomart image is:
+---[ECDSA 256]---+
|                 |
|   .             |
|    o     . .    |
| . = . . . . o   |
|. o O . S = . .  |
|.  . B  .o B o   |
|. .o. o.=.+.+ o  |
|+o+o   oE*+. o   |
|+XB.   ..==      |
+----[SHA256]-----+
  ##Copy dropbear welcome message
  inst /etc/zfsbootmenu/dracut.conf.d/banner.txt /etc/banner.txt
}
Synchronizing state of dropbear.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install disable dropbear
No initramfs generator specified; using dracut
## No version found in path /boot/vmlinuz, using 5.19.0-31-generic from kernel strings
## Identified version 5.19.0-31-generic for kernel /boot/vmlinuz
## Identified version 5.19.0-31-generic for kernel /boot/vmlinuz-5.19.0-31-generic
## No version found in path /boot/vmlinuz.old, using 5.19.0-31-generic from kernel strings
## Identified version 5.19.0-31-generic for kernel /boot/vmlinuz.old
## Latest kernel: 5.19.0-31-generic
## No version found in path /boot/vmlinuz.old, using 5.19.0-31-generic from kernel strings
## Executing: dracut -f --confdir /etc/zfsbootmenu/dracut.conf.d /tmp/oKx5Z_4VYc/zfsbootmenu.img 5.19.0-31-generic
Creating ZFSBootMenu 2.1.0 from kernel /boot/vmlinuz.old
dracut: Executing: /usr/bin/dracut -f --confdir /etc/zfsbootmenu/dracut.conf.d /tmp/oKx5Z_4VYc/zfsbootmenu.img 5.19.0-31-generic
dracut: dracut module 'mksh' will not be installed, because command 'mksh' could not be found!
dracut: dracut module 'systemd-coredump' will not be installed, because command 'coredumpctl' could not be found!
dracut: dracut module 'systemd-coredump' will not be installed, because command '/usr/lib/systemd/systemd-coredump' could not be found!
dracut: dracut module 'busybox' will not be installed, because command 'busybox' could not be found!
dracut: dracut module 'dbus-broker' will not be installed, because command 'dbus-broker' could not be found!
dracut: dracut module 'rngd' will not be installed, because command 'rngd' could not be found!
dracut: dracut module 'network-wicked' will not be installed, because command 'wicked' could not be found!
dracut: dracut module 'dbus-daemon' depends on 'systemd', which can't be installed
dracut: dracut module 'dbus' depends on 'dbus-daemon', which can't be installed
dracut: dracut module 'systemd-networkd' depends on 'dbus', which can't be installed
dracut: dracut module 'network' depends on 'systemd-networkd', which can't be installed
dracut: dracut module 'crypt-ssh' depends on 'network', which can't be installed
dracut: dracut module 'dmraid' will not be installed, because command 'dmraid' could not be found!
dracut: dracut module 'pcsc' will not be installed, because command 'pcscd' could not be found!
dracut: dracut module 'tpm2-tss' will not be installed, because command 'tpm2' could not be found!
dracut: dracut module 'cifs' depends on 'network', which can't be installed
dracut: dracut module 'fcoe' will not be installed, because command 'dcbtool' could not be found!
dracut: dracut module 'fcoe' will not be installed, because command 'fipvlan' could not be found!
dracut: dracut module 'fcoe' will not be installed, because command 'lldpad' could not be found!
dracut: dracut module 'fcoe' will not be installed, because command 'fcoemon' could not be found!
dracut: dracut module 'fcoe' will not be installed, because command 'fcoeadm' could not be found!
dracut: dracut module 'fcoe-uefi' will not be installed, because command 'dcbtool' could not be found!
dracut: dracut module 'fcoe-uefi' will not be installed, because command 'fipvlan' could not be found!
dracut: dracut module 'fcoe-uefi' will not be installed, because command 'lldpad' could not be found!
dracut: dracut module 'iscsi' depends on 'network', which can't be installed
dracut: dracut module 'nbd' depends on 'network', which can't be installed
dracut: dracut module 'nfs' depends on 'network', which can't be installed
dracut: dracut module 'nvmf' will not be installed, because command 'nvme' could not be found!
dracut: dracut module 'biosdevname' will not be installed, because command 'biosdevname' could not be found!
dracut: dracut module 'memstrack' will not be installed, because command 'memstrack' could not be found!
dracut: memstrack is not available
dracut: If you need to use rd.memdebug>=4, please install memstrack and procps-ng
dracut: dracut module 'crypt-ssh' cannot be found or installed.
Failed to create /tmp/oKx5Z_4VYc/zfsbootmenu.img

It seems to be missing all the dracut modules for reasons I'm unclear on.

Ubuntu 22.04.1 live iso, did desktop install.

Script stuck at Ign:26 https://dal.mirrors.clouvider...

The config in /etc/netplan/*.yaml needs to be deleted and netplan apply needs to be executed when a desktop is installed, else the network config becomes a unstable because it conflicts with NetworkManager and the Desktop

Hello,

Thank you for you work with this script!

I'm almost there but I cannot get my key to be accepted by the remote access setup, if I:

• add my public key to /home/<user>/.ssh/authorized_keys
• run sudo generate-zbm
• reboot
• try to login on port 222
• i still get access denied (public key) when logging into to the remote unlock ssh server

I can login directly to the system using the key without any password so I assume that the key in authorized_keys is properly added.

Some shell output:

Login and update

❯ ssh -i ~/.ssh/my_private_key [email protected]
Enter passphrase for key '<redacted>':
Welcome to Ubuntu 22.04.1 LTS (GNU/Linux 5.15.0-58-generic x86_64)

 * Documentation:  https://help.ubuntu.com
 * Management:     https://landscape.canonical.com
 * Support:        https://ubuntu.com/advantage

  System information as of Sun 15 Jan 18:33:52 CET 2023

  System load:    0.0       Temperature:                35.0 C
  Usage of /home: unknown   Processes:                  229
  Memory usage:   1%        Users logged in:            1
  Swap usage:     0%        IPv4 address for enp0s31f6: 192.168.1.26


0 updates can be applied immediately.


Last login: Sun Jan 15 18:34:29 2023 from 192.168.1.249
user@machine:~$ sudo generate-zbm
[sudo] password for user:
No initramfs generator specified; using dracut
Creating ZFSBootMenu 2.1.0 from kernel /boot/vmlinuz.old
SPL (7b85bf2c) and system (457d327f) hostids do not match!
Created backup /boot/efi/EFI/ubuntu/vmlinuz.old-bootmenu -> /boot/efi/EFI/ubuntu/vmlinuz.old-bootmenu-backup
Created backup /boot/efi/EFI/ubuntu/initramfs-bootmenu.img -> /boot/efi/EFI/ubuntu/initramfs-bootmenu-backup.img
Created initramfs image /boot/efi/EFI/ubuntu/initramfs-bootmenu.img
Created kernel image /boot/efi/EFI/ubuntu/vmlinuz.old-bootmenu

Try to access remote unlock shell

❯ ssh -i ~/.ssh/my_private_key -p 222 [email protected]
Welcome to the ZFSBootMenu initramfs shell. Enter "zbm" to start ZFSBootMenu.
Enter passphrase for key '~/.ssh/my_private_key`:
[email protected]: Permission denied (publickey).

What am I missing?

Is there any way to modify the script so that it doesn't create a swap partition?

I like Kubuntu but I don't like not being able to use the lastest KDE release.

I'm reviewing your script and trying to understand in a multi-disk setup, how are you populating the -part1 partition on each disk?

I can see within systemsetupFunc_part3():

mkdosfs -F 32 -s 1 -n EFI /dev/disk/by-id/"$DISKID"-part1 
	sleep 2
	blkid_part1=""
	blkid_part1="$(blkid -s UUID -o value /dev/disk/by-id/"${DISKID}"-part1)"
	echo "$blkid_part1"

I don't see where this is in a loop that would update the -part1 on each disk being used.

Within the initialinstall() the systemsetupFunc_part3 #Format EFI partition. is only called once, I don't see a loop.

It seem like when the one drive added to fstab dies / is removed from the system, the system will be unbootable as the the other disks do not have a populated -part1 partition and they are not referenced within fstab.

Nice work, clean script.

Goal is to install a 22.04 server with a mirror root.

As the server has not arrived yet, I try to do a dry run in a kvm.

Starting with 21.04, 21.10 and the new 22.04 ubuntu live image, the vm always end up in the uefi shell after first boot.

This is the uefi part of its configuration:

  <os>
    <type arch="x86_64" machine="pc-q35-6.0">hvm</type>
    <loader readonly="yes" type="pflash">/usr/share/OVMF/OVMF_CODE_4M.ms.fd</loader>
    <nvram>/var/lib/libvirt/qemu/nvram/test_VARS.fd</nvram>
  </os>

Here are the values I changed:

ubuntuver="jammy" #Ubuntu release to install. "hirsute" (21.04). "impish" (21.10). "jammy" (22.04).
distro_variant="server" #Ubuntu variant to install. "server" (Ubuntu server; cli only.) "desktop" (Default Ubuntu desktop install). "kubuntu" (KDE plasma desktop variant). "xubuntu" (Xfce desktop variant). "MATE" (MATE desktop variant).
user="test" #Username for new install.
PASSWORD="testtest" #Password for user in new install.
hostname="testserver" #Name to identify the new install on the network. An underscore is DNS non-compliant.
zfspassword="testtest" #Password for root pool and data pool. Minimum 8 characters.

topology_root="mirror" #single, mirror, raidz1, raidz2, or raidz3 topology on root pool.
disks_root="2" #Number of disks in array for root pool. Not used with single topology.

What am I doing wrong?

Hello and thanks for building this project.

I've had to Ctrl+C because I noticed in the middle of the run that I'd misconfigured something. When restarting I got an error stating that rpool was mounted. I tried zpool export rpool and the system said that rpool is busy. lsof shows no files using the directory, but zfs umount -a also says that it's busy. The only option I found was rebooting. Any ideas?

Tested your script from a Ubuntu 22.04.1 ISO. Update the variables, run the script, gets to the end. Reboot and it as if no ESP partitions are found. Goes right into my BIOS. I tied with Secure Boot enabled and disabled same results.

The reason I tried your script is because, I got the exact same results with my own scripts (which works fine in Virtual Box).

So I was glad I could reproduce the problem. But I have no idea how to troubleshoot this. I'm not sure if this is a rEFInd issue or not. I was hoping you had some thoughts based on your personal script testing. There is no error message or any indication of it trying. It seems like ESP partitions created do not meet what the UEFI BIOS is looking for. If I press F7 for the boot menu, my only option is to enter setup, it finds nothing unless the USB is inserted... see's that fine.

FYI - I also tested using the standard Ubuntu installer and selected ZFS with encryption. And that booted just fine. So the older ZFS on Root with pool/rpool seems to be fine. So I suspect it is related to rEFInd, but I can't come up with an intelligent question to ask.

I have 2 brand new PCs - Minisform UM560 AMD Ryzen 5 5625U 6 core 12GB RAM, 1TB NVMe (will be small home servers). Same results on both of them. Brand new AMI BIOS dated from July 2022. Nothing to update.

Hi,

I am testing the script in a VM, trying to install Kubuntu 23.04 (lucid). Latest comments suggest this should work, even if 22.04 is the main tested target.
However, I get an error when debbootstrap kicks in: E: Release signed by unknown key (key id 40976EAF437D05B5)

Here is the full log:

Setting up zfsutils-linux (2.1.9-2ubuntu1.1) ...
zfs-import-scan.service is a disabled or a static unit, not starting it.
Setting up zfs-initramfs (2.1.9-2ubuntu1.1) ...
Setting up zfs-zed (2.1.9-2ubuntu1.1) ...
Processing triggers for libc-bin (2.37-0ubuntu2) ...
Processing triggers for man-db (2.11.2-1) ...
Processing triggers for initramfs-tools (0.142ubuntu2) ...
update-initramfs is disabled since running on read-only media
Clearing partition table on disk ata-VBOX_HARDDISK_VB32f867bd-189ef66f.
Creating new GPT entries in memory.
GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.
Clearing partition table on disk ata-VBOX_HARDDISK_VBfa27f83c-617632d5.
Creating new GPT entries in memory.
GPT data structures destroyed! You may now partition the disk using fdisk or
other utilities.
Creating partitions on disk ata-VBOX_HARDDISK_VB32f867bd-189ef66f.
Creating new GPT entries in memory.
The operation has completed successfully.
The operation has completed successfully.
The operation has completed successfully.
Creating partitions on disk ata-VBOX_HARDDISK_VBfa27f83c-617632d5.
Creating new GPT entries in memory.
The operation has completed successfully.
The operation has completed successfully.
The operation has completed successfully.
I: Retrieving InRelease
I: Retrieving Release
I: Retrieving Release.gpg
I: Checking Release signature
E: Release signed by unknown key (key id 40976EAF437D05B5)
The specified keyring /usr/share/keyrings/ubuntu-archive-keyring.gpg may be incorrect or out of date.
You can find the latest Debian release key at https://ftp-master.debian.org/keys.html

Hello,
when i try to use the script to install ubuntu 20.04 i get a error, "efibootmgr : command not found "
I'm trying to install on a hp gen 10 prolient dl360

you have access to my log and the setup file.

EDIT :
needed to install the package : sudo apt-get install efibootmgr

I created a VM on my Proxmox machine and loaded the Ubuntu 22.04 desktop ISO. I installed git, set the variables as instructed and ran the initial portion of the script. However I could not choose the virtio-backed drive because it doesn't show up in /dev/disk/by-id, it only appears as /dev/vda (additional drives would be /dev/vdb and so on).

Awesome script! The only problem I've been running into is that swapon -a has been failing at line 1179 saying no such file found. I can indeed find this file after the script errors out. I tried placing a 'sleep 2' between echo /dev/disk/by-uuid/${blkid_part2} none swap defaults 0 0 >> /etc/fstab and swapon -a and this has worked flawlessly everytime. It could probably be less than 2, I chose a number for my testing.

This is running in Proxmox 7.4 on a Ryzen 9 7950x. When the vm is limited to about 12 threads, it's hit or miss if the swapon fails, mostly failing. With all 32 threads assessable to the vm, it failed 15 times in a row. With the 'sleep 2' it's worked 6 times so far.

Very minor issue, but wanted to let you know. Thank you very much for the work on this, it's been a lifesaver!

I think it would make the kernel update process easier or less error-prone. Would it be possible to remove the zfs-dkms package?

Hi! I found your script and am grateful that somebody's taken the time to build one... better than I could have!

I was wondering if you had thoughts on how best to add support for dual booting windows.

I was going to fork the script or just make local changes as needed to the partitioning, but it occurred to me that it might make sense to adjust the script slightly to use something like source to load in user specified partitioning for a more flexible approach.

Basically, the approach I was considering was changing partitionsFunc() ever so slightly:

        ALT_PARTITION_CMDS=./dualboot_win_partitions.sh
        
        while IFS= read -r diskidnum;
        do
            if [[ -f "$ALT_PARTITION_CMDS" ]]; then
                echo "Using alternative partitioning from '$ALT_PARTITION_CMDS'."
                source $ALT_PARTITION_CMDS
            else

                echo "Creating partitions on disk ${diskidnum}."
                ##2.3 create bootloader partition
                sgdisk -n1:1M:+"$EFI_boot_size"M -t1:EF00 /dev/disk/by-id/"${diskidnum}"
            
                ##2.4 create swap partition 
                ##bug with swap on zfs zvol so use swap on partition:
                ##https://github.com/zfsonlinux/zfs/issues/7734
                ##hibernate needs swap at least same size as RAM
                ##hibernate only works with unencrypted installs
                sgdisk -n2:0:+"$swap_size"M -t2:"$swap_hex_code" /dev/disk/by-id/"${diskidnum}"
            
                ##2.6 Create root pool partition
                ##Unencrypted or ZFS native encryption:
                sgdisk     -n3:0:0      -t3:BF00 /dev/disk/by-id/"${diskidnum}"
            fi

        done < /tmp/diskid_check_"${pool}".txt

And the content of dualboot_win_partitions.sh would be along the lines of:

echo "Creating partitions on disk ${diskidnum}."

##
# We use similar partition scheme to the 'default' script but instead of letting the 3rd partition (zfs / root dataset) occupy the 
#   rest of the disk, we carve out some space at the end for windows. Specifically, windows needs - as far as I can tell - two
#   partitions in addition to the EFI/Boot partition.
##

##2.3 create bootloader partition
sgdisk --new=1:1M:+"$EFI_boot_size"M -t1:EF00 /dev/disk/by-id/"${diskidnum}"
            
##2.4 create swap partition 
sgdisk --new=2:0:+"$swap_size"M -t2:"$swap_hex_code" /dev/disk/by-id/"${diskidnum}"
            
##2.6 Create root pool partition for *nix on ZFS, leaving the last 300 gigs for windows
sgdisk     --new=3:0:-300GiB      -t3:BF00 /dev/disk/by-id/"${diskidnum}"

## Create windows primary partition
sgdisk     --new=4:0:-2GiB      -t4:0700 /dev/disk/by-id/"${diskidnum}"

## Create windows recovery partition
sgdisk     --new=4:0:0      -t4:0700 /dev/disk/by-id/"${diskidnum}"

Is this an acceptable approach or is there a better/different way that would play nicely with your script?

The script ended with:

Test Summary Report

t/32.cyclic-refs.t (Wstat: 512 Tests: 0 Failed: 0)
Non-zero exit status: 2
Parse errors: No plan found in TAP output
Files=43, Tests=7434, 15 wallclock secs ( 1.51 usr 0.16 sys + 11.93 cusr 1.08 csys = 14.68 CPU)
Result: FAIL
Failed 1/43 test programs. 0/7434 subtests failed.
make: *** [Makefile:1007: test_dynamic] Error 255
TINITA/YAML-PP-0.034.tar.gz
one dependency not OK (Test::Warn); additionally test harness failed
/usr/bin/make test -- NOT OK
//hint// to see the cpan-testers results for installing this module, try:
reports TINITA/YAML-PP-0.034.tar.gz

• rm -rf /root/.cpan/build
• update-initramfs -k all -c
  perl: warning: Setting locale failed.
  perl: warning: Please check that your locale settings:
  LANGUAGE = (unset),
  LC_ALL = (unset),
  LC_TIME = "de_DE.UTF-8",
  LC_MONETARY = "de_DE.UTF-8",
  LC_ADDRESS = "de_DE.UTF-8",
  LC_TELEPHONE = "de_DE.UTF-8",
  LC_NAME = "de_DE.UTF-8",
  LC_MEASUREMENT = "de_DE.UTF-8",
  LC_IDENTIFICATION = "de_DE.UTF-8",
  LC_NUMERIC = "de_DE.UTF-8",
  LC_PAPER = "de_DE.UTF-8",
  LANG = "de_DE.UTF-8"
  are supported and installed on your system.
  perl: warning: Falling back to the standard locale ("C").
  update-initramfs: Generating /boot/initrd.img-5.15.0-46-generic
  W: Possible missing firmware /lib/firmware/ast_dp501_fw.bin for module ast
• generate-zbm
  perl: warning: Setting locale failed.
  perl: warning: Please check that your locale settings:
  LANGUAGE = (unset),
  LC_ALL = (unset),
  LC_ADDRESS = "de_DE.UTF-8",
  LC_NAME = "de_DE.UTF-8",
  LC_MONETARY = "de_DE.UTF-8",
  LC_PAPER = "de_DE.UTF-8",
  LC_IDENTIFICATION = "de_DE.UTF-8",
  LC_TELEPHONE = "de_DE.UTF-8",
  LC_MEASUREMENT = "de_DE.UTF-8",
  LC_TIME = "de_DE.UTF-8",
  LC_NUMERIC = "de_DE.UTF-8",
  LANG = "de_DE.UTF-8"
  are supported and installed on your system.
  perl: warning: Falling back to the standard locale ("C").
  Can't locate YAML/PP.pm in @inc (you may need to install the YAML::PP module) (@inc contains: /etc/perl /usr/local/lib/x86_64-linux-gnu/perl/5.34.0 /usr/local/share/perl/5.34.0 /usr/lib/x86_64-linux-gnu/perl5/5.34 /usr/share/perl5 /usr/lib/x86_64-linux-gnu/perl-base /usr/lib/x86_64-linux-gnu/perl/5.34 /usr/share/perl/5.34 /usr/local/lib/site_perl) at /usr/bin/generate-zbm line 27.
  BEGIN failed--compilation aborted at /usr/bin/generate-zbm line 27.
• cat
  main: line 73: /boot/efi/EFI/ubuntu/refind_linux.conf: No such file or directory
• '[' no = no ']'
• sed -i 's,ro quiet,ro,' /boot/efi/EFI/ubuntu/refind_linux.conf
  sed: can't read /boot/efi/EFI/ubuntu/refind_linux.conf: No such file or directory

Hi, Everyone, Thanks for the excellent Ubuntu-Server-ZFSbootmenu scripts. Suggest adapting these Ubuntu scripts to Arch Linux (and Manjaro etc). Perhaps, forking and modifying the Sithuk scrips with help from the Kayvlim examples. Is there anyone in the Arch Linux community who knows how to do this? Would this benefit the community to keep up with Ubuntu?

https://github.com/Sithuk/ubuntu-server-zfsbootmenu

See this link for information and syntax about Arch Linux root install on ZFS...
https://wiki.archlinux.org/title/User:Kayvlim/Install_UEFI_and_BIOS_compatible_Arch_Linux_with_Encrypted_ZFS_and_ZFSBootMenu#Swap...

For reference, another effort at root on ZFS install scripts that is excellent, although not as developed as the Ubuntu scripts...
https://github.com/stevleibelt/arch-linux-configuration

Replies are welcome. Thanks

Using Ubuntu 22.04.1 Server Installer.
Error:
sed: -e expression #1, char 2: unterminated address regex

How we fixed it:
ubuntu_original="$(grep -v 'security|cdrom' /etc/apt/sources.list | awk 'NF>=2{ print $2 }' | sort -u)"

How do I connect to WiFi after rebooting to do the postinstall? I am installing Kubuntu, but it seems that NetworkManager is installed/configured only after installing the distro, and I need WiFi to install the distro itself.