Personal notes and occasional posts

And stickers

I am not a person that is good at wrapping messages up in nice words and persuasive language, especially when it comes to something about money (I would never be a good salesperson). But we want to be transparent with out stuff and what we do, so that require us to write about it too. I hate begging for $, and would much rather finance everything from my own pocket but that is not really possible and would also not be that good for our users.

Running Linux.Pizza is not free – infact it costs several hundreds of dollar every year (close to $1000/year) not including the time we spend on maintenance and support. (A breakdown of the cost is further down)

Also, Linux.Pizza would really love to add new services and functionality to its portfolio, but due to limited economical resources we are not really able to make that a priority – but with your help we can change that!

LinuxPizza got donations worth $200 during 2019 and it covered the cost of operations for 2.5 month – we are extremely thankful for the generosity of our users!

So today, we are launching a small campaign to encourage everyone who have the means and want to donate to just that. And as a small thank you, we will ship you a couple of sticker that you can stick anywhere you want! For example: your laptop, your car or parents car or your stickerwall (everyone has one right?).

So in order to keep it realistic, everyone that donates atleast $10 is eligible for a small “sticker-pack” as shown in the pictures. Just let us know to whom we should ship the stickers to after the donation is made! If you have the means and will – head over to the liberapay-page or paypal-page. If you are a Brave user, you can always send a tip :) Send us an email when you have sent at donation with your email and address.

100% of the donations will go back to the Linux.Pizza project and nothing else.

Linux.Pizza Stickers! NOTE: You will receive a couple of stickers from each pile stickers!

Breakdown of the cost per year

  1. – $36
  2. – $12
  1. DNS (this includes FreeDNS environment – $144
  2. Pixelfed instance – $60
  3. Mastodon instance – $240
  4. Temporary email service – $110
  5. Power consumption and Internet connection – $150
  6. Mirror for various distros and software –$300
    • Most of the cost is sponsored by operationtulip)
  7. CDN for the Mastodon instance – $20

Some of the stuff that we would like to get started with:

  1. Nextcloud
    • This would require us to get more storage like harddrives or SSD's.
  2. Email service
    • Technically, we could rent a cheap VPS at some provider at get started. But that would certainly be hard due to the fact that GAFAM is marking mail as spam unless they come from a clean network. And cleaner networks/ISPs tend to cost more.
  3. Peertube?
    • We have gotten the question a couple of times, but we are unsure how it would fit into Linux.Pizzas services. This would require more storage anyway.

I have a custom application that my wife wrote for one of her personal projects. It turns out that the application crashes after 50-70 hours of uptime and both of us does not have the time or knowledge yet to debug that.

And that application is not that important either, it is just a website that displays various articles and pictures.

So in order to just push the problem under the rug, I just configured the system to restart the application ever 4th hour.

First, I create a service that we name “custom-application-restart“:

vi /etc/systemd/system/custom-application-restart.service
Description=restart custom application

ExecStart=/bin/systemctl restart custom-application

Next, we have to add a timer-service, note that the name of the timer-service must be the exact name of the restart-service, except that we swap out “service” to “timer”:

vi /etc/systemd/system/custom-application-restart.timer


Now, you should do the following:

systemctl daemon-reload
systemctl enable custom-application-restart.timer
systemctl start custom-application-restart.timer

Now, you should see your newly added timer-service in this list:

systemctl list-timers --all
Mon 2020-02-17 11:01:42 UTC  50min left Mon 2020-02-17 07:01:42 UTC  3h 9min ago  custom-application-restart.timer    custom-application-restart.service

Galera is a part of MariaDB and enables active/active/active replikation of databases between servers. While it necessarily dont provide any performance gains, it instead enabled a HA for the databases.

This guide assumes that you run Debian 10, which comes with MariaDB 10.3

Install MariaDB 10.3

    apt-get update
    apt-get install mariadb-server galera


It is always STRONGLY recommended to run an odd number of nodes, and atleast three nodes. This is to avoid split-brain and alot of headache and frustration in the future. Please, just set up three nodes and dont bother with a 2 node cluster.

Sure, more servers = the slower the writes will be. So it is recommended to go with atleast 3 nodes, and maximum

Galera configuration

In order to create our galera-cluster, we have to create the following file: /etc/mysql/conf.d/galera.cnf, add the following content. Just be sure that you edit it to fit your needs.


You might want to edit the “listen” address for the MariaDB installation, it is usually found in /etc/mysql/mariadb.cnf.

Configure the other servers accordingly, and execute systemctl restart mariadb-server on all nodes. You might want to execute galera_new_cluster on one of the nodes and restart all the nodes again.

Now, you can try to create a database on one node:

    create database testdb

And you should be able to see it from the other nodes:

    show databases;

And also moved over to a tiling windows manager

And you are maybe asking – why?

I once bought a keyboard together with my brother for 20 years ago, we saved our money to buy it and we did! Sadly, I do not remember the model or the brand of the keyboard – but atleast I know what it looks like! That was the ONLY keyboard I have ever bought, the rest is just from “piles of trash” mean for recycling.

About 7 years ago, I joined a random mumble channel and out of the blue we started to talk about keyboards. I've always used crappy keyboards in my life (those you find in school, libraries etc) and I did not know anything else.

I threw me back in time to when PC's where much more simpler, a time where the 5 year old me spend in front of Windows 95 and the typical 90's white colored PC, keyboard, mouse, speaker and screen. That color does reminds me of times when I did not understand the world and therefore the world was a much nicer place, for me atleast.

They keyboards where immediately thrown on my wishlist. And I dreamt of the day I could get one. I did not buy it for another 7 years – not because I did not want any, but that I instead prioritized real life stuff instead.

But finally, I did buy it. And after 10 days of waiting for it to ship from Lexington, Kentucky (US) to me here in Sweden I did get it delivered to my work. And I was the only one that was excited! Package delivered Package unpacked

I unpacked it at work, but I did not leave it there. Because this bad boy is going to be at my desk at home. There is no other place I would place such a gem somewhere else.

Keyboard on desk with two other crappy ones

As I wrote before, I've only been using crappy keyboard for my entire life, so I guess that I have to adapt to a “new way” of typing.

Here is some pictures of it: Closeup of keycaps

Day 1

I connected the keyboard to my Trisquel PC and suddenly realized that there is no built-in settings for a keyboard like this in the system. There is only a “Generic 105-key keyboard) and some other brands. I tried to find anything online regarding this keyboard and using it on Linux like inputrc configs, but so far – nothing.

Day 2

I figure that I need to write a custom .inputdrc file in order to make cool stuff of all the keys. I have already binded some keys to application specific tasks, but I would like to use my “PANIC” button that I bought extra. And I also realized that I dont have any tool for keycap removal...


Day 3

I just adapted to the keyboard fairly quickly, and I got a little addicted to the clicking of the keys. Some of the F13 – F24 keys has been mapped to application specific tasks, such as controlling the look of the terminalemulator tilda and that works perfect. I also decide that I will move to either i3 or sway, and I realized after 3 minutes that i3 is the only option of those two because Trisquel is not using wayland. So I just followed this guide and installed i3 with gaps (that's something kids like these days).

I do wonder how I will adapt to this layout thou.


Day 4

I launched i3 for the first time. And immediately started looking for a cheat sheet so I could start using it. I do liked the concept of it, and I will try to use it more and more.

I am starting to get used to the new keyboard, some keys are a little bit bigger than on normal keyboards but that's ok. My wife told me that could hear me from the other side of the apartment when I was typing – and that's not strange because this keyboard is loud, very loud.

The overall look of the keyboard is nice too, with two red buttons (ESC and a PANIC button). I do have some more keys that I want to swap out too but I lack the tools to do it.

Sadly, the Linux specific keys that I bought does not fit. So I guess I have to save that for my next Unicomp.

Day 5

Well, I've been using this keyboard for 5 days now and I do not regret it that I had to pay $105+$25+$79+$20+$40 (keyboard, keycaps, shipping, PostNord fee, import tax). It is actually one of the most expensive things I personally own (except for the car).

The typing experience is great! The noise it does is addicting/awesome and I really LOVE the retro look, especially the white color that was standard when I was a kid.

Also, to you that like it – here is video with some typing noises. Enjoy!

For those who would like to buy the real Model M,

You can check out Unicomps website, I used the Keyboard configuration tool to get mine in a Swedish layout (you have to pick Swedish Finnish).

In the last couple of month, I have not been able to sync my pretty huge library in Nextcloud due to a bug that is being ignored in the Nextcloud client. But I still need to sync my files – what to do?

It has been a source of frustration to not be able to sync down my files during this fall, so I started to look at other solutions instead, like DavFS2. With DavFS2 you can mount your Nextcloud account like you mount a HDD/SSD ora USB memory – simple!

First of all, you need to install the davfs2 package, on debian derivatives such as Ubuntu, MXLinux or PureOS:

sudo apt-get install davfs2


sudo yum install davfs2

On Fedora:

sudo dnf install davfs2

On SuSE:

sudo zypper install davfs2

Next, we want to modify the /etc/fstab file so the Nextcloud account will be mounted at boot, just modify the command provided to match your own setup:

echo " /mnt/nextcloud davfs _netdev,noauto,user,uid=USER,gid=GROUP 0 0" >> /etc/fstab

Also, I do strongly assume that you do not want to enter your login everytime you boot.

echo "/mnt/nextcloud NEXTCLOUDUSER NEXTCLOUDPASSWORD" >> /etc/davfs2/sercets

Let's finish this with adding your user into the davfs2 group

sudo usermod -a -G davfs2 USERNAME

Now, you should be able to mount it:

mount /mnt/nextcloud

Optional – encrypt transparently with gocryptfs

First, install the gocryptfs package:

sudo apt-get install gocryptfs


sudo yum install gocryptfs

On Fedora:

sudo dnf install gocryptfs

On SuSE:

`sudo zypper install gocryptfs

In this case, we are just going to create a catalogue in the homecatalogue, add is as a “plain” catalogue where the files is going to be shown decrypted for you and mount it against /mnt/nextcloud – so the files is going to be stored encrypted.

    mkdir -p ~/nextcloud_encrypted
    gocryptfs -init /mnt/nextcloud
    gocryptfs /mnt/nextcloud nextcloud_encrypted

Now, you can create a file in ~/nextcloud_encrypted, and it will show up as encrypted in the /mnt/nextcloud catalogue.

Happy sharing!

ok, here it goes: the obligatory post that everyone writes in the beginning of a new year.

2019 was a year of progress for Linux.Pizza and growth, it has been fun to curate for the small “platform” and the userbase residing here.

2020 is here, and I'll try to summarize what we hope to achieve during this year:

  • Build are more robust infrastructure: This is something that we started with last summer with OperationTulip (basically a hosted Nextcloud provider) and we will continue to do so during this year. I (Jonathan) is a part of the OperationTulip team and have contributed with alot of time and hardware into the project, so it is only fair that Linux.Pizza can reside in that environment too!

  • Get more supporters: One of the blockers of many project is money and that is mostly storage and fees for domain names and such. We have been able to mitigate a big chunk of expenses by hosting our own authorive DNS-environment and getting a room in an already existing email server. As I mentioned in the 2019 recap we have gotten donation that have been covering the cost of operation for month – and we are not able to show enough gratitude for the generosity of the people that has donated to Linux.Pizza – Thank you!!

  • Spread the word about the Librehosters: Linux.Pizza is a part of the Librehosters – a loose community of people and organizations that hosts and provide ethical and libre services for public use. We would really like to see it grow and more projects that joins the community

  • Participate in real-life events: Altough I can't attend events like FOSDEM this year, we do hope that we can join the FOSS NORTH event this spring together with the team from OperationTulip. And who knows, we may have stickers for you then? ;)

  • Linux.Pizza staff: Currently, when I am basically the sole admin of Linux.Pizza. Sure, my wife has full access to the services and will act if something would happend to me, but that's about it. So if you want to help out, or know someone that wants to help out – just contact me on Mastodon or Matrix (

Well, there is goes. Have a great day everybody!

2019 has been a great year for Linux.Pizza, we have seen hundreds of new users on our Mastodon-instance, thousands of visitors on our trashmail service and hundreds of thousand of machines using our mirror

but sadly, due to lack of time and money, I have been forced to close down a few service like Pleroma, Social relay, & Invidious.

Linux.Pizza is one of many providers in the librehost network – everyone aims to offer ethical services that focus on privacy and fairness. 2019 has been a year of growth for the network and Linux.Pizza is not excluded.

Linux.Pizza was able to achieve the following:

2020 is approaching rapidly, and Linux.Pizza is ready! If you have any suggestions of services that you would like to see, please let us know! We are always exploring ways extend our list of services that is beneficial to our users!

In your haproxy.cfg, you should do this:

    log /dev/log local0 notice
    user haproxy
    group haproxy
    stats socket /var/run/haproxy.sock mode 660 level admin
    log global
    retries 2
    timeout connect 3000
    timeout server 5000
    timeout client 5000
listen galera
    bind /var/run/mysqld/mysqld.sock mode 666
    option mysql-check user HAPROXY
    balance first
    server galera1 check maxconn 64
    server galera2 check maxconn 64
    server galera3 check maxconn 64

And why do I listen on a socket? That is because an application will assume that mysql/mariadb are listening on a socket when you specify “localhost” by default.

This assumes that you are running your Linuxsystem as a virtual server on VMWare, KVM, XEN or AHV.

We have been there a couple of times – your application need more RAM. Or just want more RAM in order to have your Minecraft server run smoothly.

  1. First of all, add the amount of RAM to your VM via your hypervisor.
  2. Second, we need to tell the kernel that there is more RAM available for use, we can do that by using this simple script:

for f in /sys/devices/system/memory/memory*/state ; do grep -v -q online $f || continue echo -n "Bringing $f online... " echo online > $f || continue echo OK done

That's all! You can now verify with your favorite way to check RAM usage.

This assumes that you are running your Linuxsystem as a virtual server on VMWare, KVM, XEN or AHV.

Be sure to take a backup first!

It is quite common – for me atleast – that a busy databaseserver is going to use more and more space. And in most cases it is not always that attractive to have to restart the server and resize it with a LiveCD such as gparted live.

Luckily, there is possible to expand the partitions while your machine is running. I will try to guide you with this step to step guide on how you can successfully proceed with it:

If you are using LVM, read until 12, and skip 13.

  1. First of all, add the storage in you hypervisor.
  2. Next, you want to tell the kernel that to rescan the bulk device: echo 1 > /sys/class/block/sda/device/rescan fdisk are now able to “see” the expanded disk, in this case /dev/sda
  3. Now, we need to do the actual expansion of the partinion, so we need to “delete” the partion and add it again via fdisk – dont worry, we are not going to write the changes to disk while the partinion is deleted: fdisk /dev/sda
  4. Just to make sure, check the partinion by selecting p, if the disk looks like it should, you may continue
  5. Delete the partition by pressing d
  6. Press n to create a new partinion
  7. Press p to chose “Primary Partition”
  8. Press 1 to chose the partition number 1
  9. Press ENTERtwice
  10. And lastly, doublecheck the changes by pressing p, note the difference from the first time you did it. You should see that the partition has been expanded.
  11. !The following actions will write the changes to disk! Now, in order to make is real, press w in order to write the changes to disk, you will see a warning about that the partition table has been changed. Do not worry, it is expected.
  12. Now, run partprobe (you have to install parted for this to work). If you are using an older kernel, you maybe would need to reboot if the partprobe task did not work.
  13. Now, we just need to expand the filesystem itself in order to actually use the newly added space: resize2fs -p /dev/sda1

Verify with df -h or lsblk

Steps for if you are using LVM

I assume that you have done step 1-12 in the previous section.

  1. First, expand your physical volume: pvresize /dev/sda1
  2. Then, we will proceed with expanding the Logical volume, replace “X1” with the Logical Volume you want to extend: lvextend --extents +100%FREE /dev/vg0/X1
  3. After that is done, we have to expand the filesystem, again – replace “X1” with the Logical Volume you want to extend resize2fs /dev/vg0/X1

Verify with df -h or lsblk

You should be able to see the new size now.