Linux, Unix, and whatever they call that world these days
I have been using different *nixes for a while now. Long enough that one simply gets used to some things. For example, I still turn off my OS using init 0 and reboot it using init 6. It’s just a habit I couldn’t shake.
Well, I guess those days are over. As of Ubuntu 26.04 these old commands will now give error:
Excess argumentsWell, there go decades of muscle memory…
My general destop choice is Kubuntu. I like Ubuntu package management and general compatibility and I love Plasma. I do have other distributions around my home, but my main setup sticks to Kubuntu. So, with the new version, came time to have a new setup instructions.
Honestly, installers these days are good and you will probably be happy with default installation unless you have special needs. In my case, I do have a few quirks the default installer doesn’t support to my liking:
Since installed doesn’t do that, I do quite a few manual steps to make it works. Mind you, everything is set for copy/paste so they are not that troublesome or excessive as they look. In any case, here they are.
The very first thing to do is to boot into installation. Once in live desktop I usually setup networking and add a few packages so I can acccess it from the network.
sudo apt update ; sudo apt install openssh-server
passwdIf you want to execute these commands directly, you can freely skip those steps. If you do want to access the system over network, do not forget to disable sleep as it will happily interrupt your installation steps.
Once in terminal, we want to switch to the root user.
sudo -iThen we add a few prerequisite packages. Most notably ZFS and something to partition disk with (in my case sgdisk).
apt update
apt install -y gdisk zfsutils-linuxNow, as promised, I like to set a few variables.
DISK1=/dev/disk/by-id/<disk>
HOST=<host>
USERNAME=<user>Note I use absolute path to the disk. I found it least problematic when scripting.
With that sorted out, we’re onto setting up partitions. My partitions are:
Note this will destroy any data you might have, thus there is no way back after this.
blkdiscard -f $DISK1 2>/dev/null
sgdisk --zap-all $DISK1
sgdisk -n1:1M:+127M -t1:EF00 -c1:EFI $DISK1
sgdisk -n2:0:+1920M -t2:8300 -c2:Boot $DISK1
sgdisk -n3:0:+32G -t3:8200 -c3:Swap $DISK1
sgdisk -a 8 -I -n4:0:0 -t4:8309 -c4:LUKS $DISK1
sgdisk --print $DISK1With partitions created, we can collect their IDs.
DISK1P1=`blkid -s PARTUUID -o value $DISK1-part1`
DISK1P2=`blkid -s PARTUUID -o value $DISK1-part2`
DISK1P3=`blkid -s PARTUUID -o value $DISK1-part3`
DISK1P4=`blkid -s PARTUUID -o value $DISK1-part4`I’ve been going back and forth over the years on whether to use simpler partition names or UUIDs for the rest of setup. However, UUIDs are simply more fool-proof. For example, if you every clone your drive, it will just work. Partition names get too messy when you move disks around.
Now let’s setup LUKS encryption for ZFS.
cryptsetup luksFormat -q --type luks2 \
--sector-size 4096 \
--cipher aes-xts-plain64 --key-size 256 \
--pbkdf argon2i /dev/disk/by-partuuid/$DISK1P4While your disk is encrypted, you still need to open it. I will assume you’re gonna use SSD, thus usage of perf-no_write_workqueue and perf-no_read_workqueue options.
cryptsetup luksOpen \
--persistent --allow-discards \
--perf-no_write_workqueue --perf-no_read_workqueue \
/dev/disk/by-partuuid/$DISK1P4 $DISK1P4With this we can create our pool.
zpool create \
-o ashift=12 -o autotrim=on \
-O compression=lz4 -O normalization=formD \
-O acltype=posixacl -O xattr=sa -O dnodesize=auto -O atime=off \
-O canmount=off -O mountpoint=none -R /mnt/install \
${HOST^} /dev/mapper/$DISK1P4I also recommend setting up quota but that’s something you can do later depending on how big the disk is.
On top of ZFS pool we can now create our system dataset.
zfs create \
-o devices=on \
-o canmount=noauto -o mountpoint=/ \
${HOST^}/System
zfs mount ${HOST^}/SystemAnd a separate home dataset.
zfs create \
-o canmount=noauto -o mountpoint=/home \
${HOST^}/Home
zfs mount ${HOST^}/Home
zfs set canmount=on ${HOST^}/HomeOur boot partition wont be encrypted and that’s by design. It’s not ideal but recovering system with encrypted partition is such pain that I am willing to compromise.
yes | mkfs.ext4 /dev/disk/by-partuuid/$DISK1P2
mkdir /mnt/install/boot/
mount /dev/disk/by-partuuid/$DISK1P2 /mnt/install/boot/Our EFI partition gets its FAT32 home too.
mkfs.msdos -F 32 -n EFI -i 4d65646f /dev/disk/by-partuuid/$DISK1P1
mkdir /mnt/install/boot/efi/
mount /dev/disk/by-partuuid/$DISK1P1 /mnt/install/boot/efi/And lastly, our swap partition gets its own LUKS encrption. I strongly suggest the same password as for data partition as it will save you time typing.
cryptsetup luksFormat -q --type luks2 \
--sector-size 4096 \
--cipher aes-xts-plain64 --key-size 256 \
--pbkdf argon2i /dev/disk/by-partuuid/$DISK1P3Mind you, it too needs opening.
cryptsetup luksOpen \
--persistent --allow-discards \
--perf-no_write_workqueue --perf-no_read_workqueue \
/dev/disk/by-partuuid/$DISK1P3 $DISK1P3And formatting.
mkswap /dev/mapper/$DISK1P3At this point I like to disable IPv6. I am not sure if it’s my network or a general issue, but IPv4-only network is MUCH faster in live USB environment. Mind you, this is just a temporary thing - installed system will support IPv6 just fine.
sysctl -w net.ipv6.conf.all.disable_ipv6=1
sysctl -w net.ipv6.conf.default.disable_ipv6=1
sysctl -w net.ipv6.conf.lo.disable_ipv6=1Finally, we get to install at least minimum Resolute Racoon.
apt update
apt dist-upgrade --yes
apt install --yes debootstrap
debootstrap --components=main,restricted,universe,multiverse resolute /mnt/install/To help it going, I like to set its hostname and copy over network details. The last step is optional but it helps you preserver wireless network password.
echo $HOST > /mnt/install/etc/hostname
sed "s/kubuntu/$HOST/" /etc/hosts > /mnt/install/etc/hosts
cp /etc/netplan/*.yaml /mnt/install/etc/netplan/At this time we switch into our new 26.04 environment. Anything you do after this step permanently impacts your installation. For all practical purposes, it’s like you’re logged in your new setup. Because you are.
mount --rbind /dev /mnt/install/dev
mount --rbind /proc /mnt/install/proc
mount --rbind /sys /mnt/install/sys
chroot /mnt/install /usr/bin/env \
DISK1P1=$DISK1P1 DISK1P2=$DISK1P2 DISK1P3=$DISK1P3 DISK1P4=$DISK1P4 \
HOST=$HOST USERNAME=$USERNAME \
bash --loginI like to setup my locale.
dpkg-reconfigure localesAnd my timezone.
dpkg-reconfigure tzdataWith those out of way, it’s time to tell our system which partitions it will need to decrypt.
echo -n > /etc/crypttab
echo "$DISK1P3 PARTUUID=$DISK1P3 none luks,discard,initramfs,keyscript=decrypt_keyctl" >> /etc/crypttab
echo "$DISK1P4 PARTUUID=$DISK1P4 none luks,discard,initramfs,keyscript=decrypt_keyctl" >> /etc/crypttabThen we need to tell it where to mount what.
echo -n > /etc/fstab
echo "PARTUUID=$DISK1P2 /boot ext4 noatime,nofail,x-systemd.device-timeout=3s 0 1" >> /etc/fstab
echo "PARTUUID=$DISK1P1 /boot/efi vfat noatime,nofail,x-systemd.device-timeout=3s 0 1" >> /etc/fstab
echo "/dev/mapper/$DISK1P4 none auto nofail,nosuid,nodev,noauto 0 0" >> /etc/fstab
echo "/dev/mapper/$DISK1P3 none swap sw,nofail 0 0" >> /etc/fstabOur new kernel packages go next.
apt update
apt install --yes --no-install-recommends linux-image-generic linux-headers-genericFollowed by packages needed for making it boot.
apt install --yes zfs-initramfs cryptsetup keyutils plymouth-theme-spinner
update-initramfs -c -k allTo make EFI work, we need to install GRUB too. I modify its /etc/default/grub a bit to ensure it will resume from our swap. Note that you bootloader-id should stay ubuntu if you’re using secure boot.
apt install --yes grub-efi-amd64-signed shim-signed
sed -i "s/^GRUB_CMDLINE_LINUX_DEFAULT.*/GRUB_CMDLINE_LINUX_DEFAULT=\"quiet splash \
RESUME=UUID=$(blkid -s UUID -o value /dev/mapper/$DISK1P3)\"/" \
/etc/default/grub
sed -i 's/^GRUB_TIMEOUT=.*/GRUB_TIMEOUT=1/' /etc/default/grub
echo "GRUB_DISABLE_MEMTEST=true" >> /etc/default/grub
update-grub
grub-install --target=x86_64-efi --efi-directory=/boot/efi \
--bootloader-id=Ubuntu --recheck --no-floppyWith booting sorted, it’s time to setup desktop and I like to start by forbidding snap altogether. Yes, there is a certail level of madness in using Ubuntu without snap but that’s story for another time.
apt remove --yes snapd 2>/dev/null
echo 'Package: snapd' > /etc/apt/preferences.d/snapd
echo 'Pin: release *' >> /etc/apt/preferences.d/snapd
echo 'Pin-Priority: -1' >> /etc/apt/preferences.d/snapd
apt updateAt last, we come to the desktop installation itself. I like to add a few tools here too.
apt install --yes kubuntu-desktop man-db lshw iw wgetSince disabling snap also disabled installation of Firefox, we need to readd it from PPA to have a browser installed once everything is done.
add-apt-repository --yes ppa:mozillateam/ppa
cat << 'EOF' | sed 's/^ //' | tee /etc/apt/preferences.d/mozillateamppa
Package: firefox*
Pin: release o=LP-PPA-mozillateam
Pin-Priority: 501
EOF
apt update && apt install --yes firefoxWith desktop installed, we’re literally on the final stretch of completing install. So, it’s a good time to setup our user.
adduser --disabled-password --gecos '' $USERNAME
usermod -a -G adm,cdrom,dialout,dip,lpadmin,plugdev,sudo,tty $USERNAME
echo "$USERNAME ALL=NOPASSWD:ALL" > /etc/sudoers.d/$USERNAME
passwd $USERNAMEHere I also like to setup my sleep options in order to enable hibernation.
sed -i 's/.*AllowSuspend=.*/AllowSuspend=yes/' /etc/systemd/sleep.conf
sed -i 's/.*AllowHibernation=.*/AllowHibernation=yes/' /etc/systemd/sleep.conf
sed -i 's/.*AllowSuspendThenHibernate=.*/AllowSuspendThenHibernate=yes/' /etc/systemd/sleep.conf
sed -i 's/.*HibernateMode=.*/HibernateMode=platform shutdown/' /etc/systemd/sleep.conf
sed -i 's/.*HibernateDelaySec=.*/HibernateDelaySec=13min/' /etc/systemd/sleep.confMy preferred way of handling sleep on laptop is to suspend when closing the screen and hibernate on power key. If your preference differ, skip this step.
apt install -y pm-utils
sed -i 's/.*HandlePowerKey=.*/HandlePowerKey=hibernate/' /etc/systemd/logind.conf
sed -i 's/.*HandleLidSwitch=.*/HandleLidSwitch=suspend-then-hibernate/' /etc/systemd/logind.conf
sed -i 's/.*HandleLidSwitchExternalPower=.*/HandleLidSwitchExternalPower=ignore/' /etc/systemd/logind.conf
sed -i 's/.*HoldoffTimeoutSec=.*/HoldoffTimeoutSec=13s/' /etc/systemd/logind.confSince we’re good with our new environment we can exit it.
exitNow we can disable devices flag on the system dataset and ensure all data is written. That is followed by unmounting all partitions and exporting ZFS pool.
zfs set devices=off ${HOST^}/System
sync
umount /mnt/install/boot/efi
umount /mnt/install/boot
mount | grep -v zfs | tac | awk '/\/mnt/ {print $3}' | xargs -i{} umount -lf {}
zpool export -aWith cleanup done, we can finally reboot.
rebootIf everything went well, you now get to enjoy KDE Plasma.
My tmux setup, as many, includes remote server hostname as part of the “tab” title. As I use ssh to move from server to server, title keeps me informed where exactly I am. Except it doesn’t work with bare Alpine Linux.
The easiest way I got it working again is by adding the following lines after my last PS1 is set. For example, if using very bare setup, adding this line to ~/.profile is sufficient.
PS1='\[\ek\h:\W\]'$PS1This will prefix any PS1 output with a terminal title setup sequence (ESC k). Content here can be whatever, but I like to make it <hostname>:<dir> (\h:\W). This gives me the most information in minimal character count.
Happy tmuxing!
I still like using SpeedTest by Ookla. It’s not the only kid in the town anymore but old habits die slowly. And honestly, it is a decent speed tool.
So, when I wanted to measure Internet speed from my Ubuntu server, I was happy to see it was available. And install is trivial
sudo apt install speedtest-cliHowever, on my Ubuntu 24.04 installation this doesn’t really help since result is always 0.00. Never mind, there is a source availabe in now archived speedtest-cli repo. With a small change, this script will work a charm.
mkdir -p ~/Downloads
cd ~/Downloads
wget https://raw.githubusercontent.com/sivel/speedtest-cli/refs/heads/master/speedtest.py
chmod +x speedtest.py
sed -i '1s/python/python3/' speedtest.py
./speedtest.pyWith sudo tool getting its Rust variant, it was bound sooner or later to have an incompatibility that is in the eye of the beholder. That dubious honor fell onto pwfeedback setting.
Most complete reporting was already done by Brodie Robertson so watch that video for details. Suffice to say, I am firmly in the camp of those who believe that the new behavior is correct. Let them have asterisk!
But, if you don’t want to wait for the bright future but you want this password echo behavior now, this is really easy to achieve.
echo "Defaults pwfeedback" | sudo tee /etc/sudoers.d/pwfeedback >/dev/null
sudo chmod 440 /etc/sudoers.d/pwfeedbackNow you can experience the future today.
Running KDE is usually quite troublefree. However, on my new install I faced unusual problem. My photos and media on samba share simply would not open. I could see text files just fine. But no movie for me.
Interestingly, this behavior seemed limited to MPV and Haruna. Other media players seemed unaffected by whatever afflicted them. It took a bit of investigation but I traced the issue to the following line in their .desktop file:
X-KDE-Protocols=ftp,http,https,mms,rtmp,rtsp,sftp,smb,srt,rist,webdav,webdavsWhile they both claim to understand samba protocol (smb), removing it from supported protocols actually solved the issue. It seems that my samba server was simply not to their liking. After doing a bit of investigation, I am still puzzled why exactly - especially since it works just fine on another computer. Some dependency is missing and, while finding it would be possible, it would also waste my time when I already have solution.
Just removing smb works like a charm:
sudo sed -i 's/,smb,/,/g' /usr/share/applications/mpv.desktop
sudo sed -i 's/,smb,/,/g' /usr/share/applications/org.kde.haruna.desktopIt’s a temporary fix but Kubuntu 26.04 is just around the corner anyhow.
My “standard” partition scheme is always, EFI, Boot, Swap, Data. I have been toying with an idea to swap EFI and Boot around but I stuck with the same order for a while now. However, sizes have changed over the time. One that changed the most is EFI.
Under Linux, assuming you have a separate Boot partition, you can have EFI really small. I’ve ran it at 32 MB for a while, not observing any negative issues. Until I did.
While most of the time Linux doesn’t really use EFI partition for much, things get interesting with firmware updates (fwupdmgr update). The only way for Linux (or any other OS, for that matter) to pass files to UEFI is a common partition. That common partition can only be FAT32 EFI. And, depending on your system, those files might need over 64 MB.
Translated, your EFI needs to be able to have more than 64 MB. How much more? Playing with power of two values is an unofficial standard and thus 128 MB is the next good spot.
But, is that future proof? Considering BIOS images are 32 MB these days (256 MBit), it’ll take a while before 128 MB is not enough
I am almost certain that more space will be needed at undefined time in the future. However, I won’t lose any sleep about it for at least couple of years.
PS: If you are dual-booting Windows, I would say doubling the partition might be a good idea as Windows does have a larger footprint on EFI partition.
New BIOS for Framework 16, brought also a problem. Every boot I got a message that my keyboard firmware is outdated. It was true, but also annoying warning as it paused the boot process. Normal person would just update keyboard but I had a few customizations and thus could not do the same.
What customizations? Well…
Could I live without those? Yes, but I still would prefer not to. So, I had to redo my changes on top of v0.3.1 which was simple enough as I just pretty much cherry-picked my changes directly. Maybe for the next version I also squash a few of them but I was lazy this time.
Short flash later and my numpad had a newer firmware with the same behavior. Happily I proceeded to flash the keyboard but in my excitement, I forgot to change my command line from framework/numpad to framework/ansi. Thus, I flashed my numpad firmware onto my keyboard. Annoying mistake but easily corrected by just reflashing the correct firmware on top of it. Or so I though.
My keyboard still didn’t work even after the correct firmware has been flashed. Some keys did produce something but it was never a correct letter. Did I brick my keyboard? Well, fortunately for me, with QMK answer is never yes.
QMK flashing doesn’t erase the full EEPROM when new version is loaded. So, if you end up corrupting the whole thing by flashing something very incompatible, you cannot just flash new firmware and be ok. What you need is to erase the old firmware all together. There are a few tools to do it, but I like this one.
Once EEPROM was erased, flashing my keyboard firmware worked like a charm.
I have most of workloads on my server setup in Docker. It makes for self-documenting configuration, easier move to other machine, and less dependencies on the underlying OS. But one workload I kept on server itself - reporting.
The way my scripts were written, they required proper ZFS access. Now, I could adjust script to loopback via SSH to OS itself, but there was an easier way.
Docker allows for device exposing, and in the case of ZFS, all you need is to point container toward /dev/zfs:
devices:
- /dev/zfs:/dev/zfsNow your scripts in container will have proper ZFS access.
After messing with my remote server, there came a time for a reboot. Simple enough - but this time it ended in error.
Call to Reboot failed: Connection timed outI’ve been using Linux servers for decades now and I was never faced with this issue. Fortunately, somebody at Unix StackExchange did.
Solution was to manually enter a few letters into /proc/sysrq-trigger, one letter at a time.
echo s > /proc/sysrq-trigger ; echo u > /proc/sysrq-trigger ; echo b > /proc/sysrq-triggerThis (attempts to) execute three distinct commands:
s: syncs all file systemsu: makes all file systems read-onlyb: reboots the system in agressive manner.If you are curious about what other things you can do, kernel.org has a nice documentation page.