linux on log.Printf

how old is ann?

Sat, 01 Aug 2026 00:00:00 +0000

Mary is 24 years old.

She is twice as old as Ann was when Mary was as old as Ann is now.

How old is Ann?

It sounds simple. It probably isn’t.

logprintf is now indieweb compliant

Sat, 01 Aug 2026 00:00:00 +0000

I’ve updated logprintf.com to support IndieWeb standards.

The goal is simple: keep ownership of my content while making the site easier to connect with the wider web.

A small technical upgrade, but an important principle: own your content.

find what is using a port on linux

Fri, 31 Jul 2026 00:00:00 +0000

To find which process is using a port on Linux, run:

sudo ss -lptn 'sport = :8080'

Replace 8080 with the port you want to inspect.

You can also use:

sudo lsof -i :8080

Useful when a service refuses to start because its port is already occupied.

sudo, n8n and the missing terminal

Fri, 17 Jul 2026 00:00:00 +0000

I have an n8n workflow that SSHes into my server and runs smartctl to check drive health. It kept failing with “sudo: a terminal is required to read the password”.

The SSH node opens a non-interactive session with no TTY, so sudo has no way to prompt me. The fix is to stop it asking at all: a passwordless rule in /etc/sudoers.d, scoped to smartctl only rather than opening up sudo entirely.

That still didn’t work, because sudoers matches on the exact command path you write in the rule. On Debian /usr/sbin isn’t in a normal user’s PATH, only root’s, so which smartctl comes back empty and it’s easy to write the rule against a path that doesn’t resolve. Use /usr/sbin/smartctl in both the sudoers rule and the script and it goes through.

rsync couldn't delete what it copied

Fri, 26 Jun 2026 00:00:00 +0000

One of the drives in my MergerFS pool started showing pending sectors, so I mounted it read-only and rsynced everything off with --remove-source-files.

Two terabytes later, rsync finished the copy and then threw errors on every single delete. Obvious in hindsight: read-only was the whole point of the mount, and --remove-source-files needs write access to unlink the originals. The data was safely on the other side, it just couldn’t clean up after itself.

Remounted with mount -o remount,rw and ran the same command again. rsync skips anything already transferred, so the second pass was just the deletions and took minutes. Worth knowing before you panic at a screen full of errors.

nohup is enough

Fri, 08 May 2026 00:00:00 +0000

Starting a multi-hour rsync over SSH, I reached for tmux out of habit and got a socket error instead of a session.

I could have fixed the socket, but the requirement was smaller than the tool. All I actually needed was for the process to survive my connection dropping, which is one line: nohup rsync ... > /tmp/migration.log 2>&1 &, then tail -f on the log whenever I want to check in.

Worth knowing because it needs nothing installed, which matters exactly when you’re on a machine you’ve just started troubleshooting. tmux is better when you want to come back to a live terminal — but for fire-and-watch-the-log jobs, it’s a dependency you don’t need.

recreate a container after configuration changes

Thu, 23 Apr 2026 00:00:00 +0000

I changed a container’s environment variables and restarted it, but nothing changed.

A restart only stops and starts the existing container. It does not necessarily rebuild it with the new configuration.

Recreate the service instead:

docker compose up -d --force-recreate service_name

When configuration changes are ignored, make sure the container was actually recreated.

check file permissions when a container cannot write

Tue, 07 Apr 2026 00:00:00 +0000

A container may start correctly but fail when it tries to write to a mounted directory.

Check the directory ownership:

ls -ld /path/to/data

Then compare it with the user running inside the container:

docker exec container_name id

Many mysterious database and configuration errors are simply volume permission problems.

Two SMART logs, two different stories

Sat, 04 Apr 2026 00:00:00 +0000

A drive in my pool was failing extended self-tests while the SMART error log stayed completely clean. It reads like a contradiction and I nearly talked myself into trusting the clean log.

They track different things. The error log records command-level ATA errors the firmware explicitly reports — unrecoverable reads, write faults. The self-test log records whether the test procedure completed. A drive can abort a test early on some internal condition without ever generating a loggable error, which is exactly what “unknown failure” at 90% remaining means: it gave up almost immediately and had nothing to file.

The self-test result is the one to believe. Combined with reallocated sectors it was enough to start the migration, and no, firmware updates don’t repair physical sectors.

read the full error message

Tue, 31 Mar 2026 00:00:00 +0000

I sometimes focus on the feature that is not working and overlook the actual error message.

In one case, the application clearly said:

Expected string, got array

That was the problem.

Before searching for complicated fixes, read the complete error and compare the received value with the expected type.

how to set up static IP address

Sat, 14 Sep 2024 00:00:00 +0000

Setting up a static IP address can help ensure that a device keeps the same IP address over time, which is especially useful for servers or other network services.

Below are the steps to configure a static IP address on a Linux machine using the /etc/network/interfaces file.

First, open the terminal application on your local machine or log into your remote server using SSH. If you are working on a remote machine, use the following command to establish an SSH connection:

ssh user@remote-server-ip

Replace user with your actual username and remote-server-ip with the IP address of your remote machine.

Before making any changes, it’s a good practice to back up your existing network configuration file. Run the following command to create a backup of the /etc/network/interfaces file:

sudo cp /etc/network/interfaces /root/interfaces.bak

This backup allows you to restore the configuration in case something goes wrong.

Now, open the /etc/network/interfaces file in your preferred text editor to make changes. I use vim:

sudo vim /etc/network/interfaces

Find the section of the file that corresponds to your Ethernet interface. In this example, we’ll configure the enp2s0 interface with the following details:


IP Address: 192.168.1.158
Subnet Mask: 255.255.255.0
Default Gateway: 192.168.1.1
DNS Nameservers: 192.168.1.1, 8.8.8.8, 1.1.1.1

Add or modify the lines in the file as follows:

auto enp2s0
iface enp2s0 inet static
    address 192.168.1.158
    netmask 255.255.255.0
    gateway 192.168.1.1
    dns-nameservers 192.168.1.1 8.8.8.8 1.1.1.1

Make sure to replace enp2s0 with the correct name of your Ethernet interface, which you can find using the ip a command.

To apply the changes, restart the networking service with the following command:

sudo systemctl restart networking.service

Make sure service restarted without any errors with the following command:

sudo systemctl status networking.service

You can verify that the static IP has been successfully configured by running:

ip a

Check that the IP address for the enp2s0 interface matches the static IP you set (192.168.1.158 in this example).

how to set up static IP address on freebsd machine

Sat, 14 Sep 2024 00:00:00 +0000

Setting up a static IP address can help ensure that a device keeps the same IP address over time, which is especially useful for servers or other network services.

Below are the steps to configure a static IP address on a FreeBSD machine using the /etc/rc.conf file.

The network configuration in FreeBSD is controlled through the /etc/rc.conf file. Begin by opening this file with your preferred text editor. In this example, I will use vim:

vim /etc/rc.conf

Locate the line in the file that configures the Ethernet interface (usually em0 or something similar). It will look something like this:

ifconfig_em0="DHCP"

This line tells FreeBSD to use DHCP for assigning an IP address. To assign a static IP, replace the line with the following:

ifconfig_em0="inet 192.168.1.61 netmask 255.255.255.0"

In this example:

192.168.1.61 is the static IP address you want to assign to your machine.
255.255.255.0 is the subnet mask.

Make sure to adjust the IP address and subnet mask according to your network configuration.

Next, we need to specify the default gateway (the router’s IP address) that your machine will use to access devices outside its local network. Add or update the following line in /etc/rc.conf:

defaultrouter="192.168.1.1"

Replace 192.168.1.1 with the actual IP address of your network’s default gateway. Save and exit.

To apply the changes, you’ll need to restart the networking service on FreeBSD. This can be done by either rebooting the system or restarting the networking service with the following command:

service netif restart

You may also want to restart the routing service to ensure the default gateway is applied:

service routing restart

Once the system is back up and running, you can verify the static IP configuration using the ifconfig command:

ifconfig em0

Check that the em0 interface is now using the static IP address (192.168.1.1 in this example) and has the correct subnet mask.

persistent permissions on /dev/dri

Tue, 10 Sep 2024 00:00:00 +0000

So in my latest post regarding setting up your media server app (Pex/Emby/Jellyfin) so that it uses your iGPU utilizing QuickSync I mentioned that you needed to change permissions on /dev/dri so that your docker container can access it. In order to do that you do:

sudo chmod -R 777 /dev/dri

So, that works. But the problem with that is, when you restart your server the permissions are reset and you need to repeat the same command again.

What you can do to make the permission changes permanent is that, you can create a udev rule. That will ensure that the correct permissions are applied to the /dev/dri device at boot time. So here’s how:

First we create a udev rule file:

sudo vim /etc/udev/rules.d/99-dri.rules

Then we add the following content to the file:

SUBSYSTEM=="drm", MODE="0777"

This rule sets the permissions of any device under the drm subsystem (like /dev/dri) to 0777.

After saving the file, reload the udev rules:

sudo udevadm control --reload-rules

Reboot your system and your docker containers can access /dev/dri persistently after each reboot:

sudo reboot

setup plex/emby/jellyfin hardware transcoding via intel igpu

Mon, 09 Sep 2024 00:00:00 +0000

So I’ve been having issues with enabling hardware transcode in my Emby Container on my Debian machine with i5-8500. These are the steps I had to take to enable hardware transcoding utilizing QuickSync.

First I had to add non-free to the debian sources in order to install Intel non-free drivers.

sudo nano /etc/apt/sources.list

I added contrib and non-free to the end of the relevant lines so my sources.list file looks like below:

#deb cdrom:[Debian GNU/Linux 12.6.0 _Bookworm_ - Official amd64 NETINST with firmware 20240629-10:18]/ bookworm contrib main non-free-firmware

deb http://deb.debian.org/debian/ bookworm main non-free-firmware contrib non-free
deb-src http://deb.debian.org/debian/ bookworm main non-free-firmware contrib non-free

deb http://security.debian.org/debian-security bookworm-security main non-free-firmware contrib non-free
deb-src http://security.debian.org/debian-security bookworm-security main non-free-firmware contrib non-free

# bookworm-updates, to get updates before a point release is made;
# see https://www.debian.org/doc/manuals/debian-reference/ch02.en.html#_updates_and_backports
deb http://deb.debian.org/debian/ bookworm-updates main non-free-firmware contrib non-free
deb-src http://deb.debian.org/debian/ bookworm-updates main non-free-firmware contrib non-free

# This system was installed using small removable media
# (e.g. netinst, live or single CD). The matching "deb cdrom"
# entries were disabled at the end of the installation process.
# For information about how to configure apt package sources,
# see the sources.list(5) manual.

Then we go ahead and update our package list:

sudo apt update

Now we install the Intel non-free drivers:

sudo apt install intel-media-va-driver-non-free

Now we make sure everything is in order with our Intel drivers:

ls /dev/dri

You should see something like this:

by-path/   card0   renderD128

If not try:

modprobe i915

And checking it out again via:

ls /dev/dri

Now you should see something like this:

by-path/   card0   renderD128

Now we need to make sure our Emby/Plex/Jellyfin docker container has access to the Intel iGPU.

First we need to set the permissions:

sudo chmod -R 777 /dev/dri

Now we edit our docker compose file so the /dev/dri:/devdri mapping is correct. My compose file looks like as below:

services:
  emby:
    image: emby/embyserver:latest
    container_name: emby
    devices:
      - /dev/dri:/dev/dri  
    environment:
      - UID=1000 # Replace with your user ID
      - GID=1000 # Replace with your group ID
    volumes:
      - ./config:/config # Replace with the path to your Emby config directory
      - /mnt/merged:/mnt/media # Replace with the path to your media files
    ports:
      - 8097:8096
      - 8920:8920 # Optional for HTTPS
    restart: unless-stopped

Now you should be able to setup your Emby/Plex/Jellyfin so that it can utilize your Intel iGPU to hardware transcode via QuickSync.

do nothing when laptop lid is closed

Thu, 05 Sep 2024 00:00:00 +0000

So if you’re like me and use one of your old laptops in your cluster as a headless server (mine is on Adguard duty) this could help you. You’d most likely want to keep your laptop lid/screen closed. If you do so and you want your Debian install to do nothing when the lid is closed all you have to do is:

edit /etc/systemd/logind

sudo nano /etc/systemd/logind

uncomment and set the following options:

HandleLidSwitch=ignore
HandleLidSwitchExternalPowe r=ignore
HandleLidSwitchDocked=ignore
LidSwitchIgnoreInhibited=no

reboot and you are set to go.

sudo reboot

docker media server setup

Tue, 03 Sep 2024 00:00:00 +0000

Below is my docker-compose.yml I use for my media server needs. Stack includes Sonarr, Radarr and Sabnzbd for download management. For streaming media it includes Emby. We tried both Plex and Jellyfin before Emby. Emby works best with our Samsung Smart TV so we stayed with Emby.


services:
  sabnzbd:
    image: linuxserver/sabnzbd
    container_name: sabnzbd
    environment:
      - PUID=1000 # Replace with your user ID
      - PGID=1000 # Replace with your group ID
    volumes:
      - ./config/sabnzbd:/config # Replace with the path to your config directory
      - ./downloads:/downloads
      - ./usenet:/usenet
      - /mnt/merged:/mnt/merged # Replace with the path to your media files
    ports:
      - 8080:8080
    restart: unless-stopped

  radarr:
    image: linuxserver/radarr
    container_name: radarr
    environment:
      - PUID=1000 # Replace with your user ID
      - PGID=1000 # Replace with your group ID
    volumes:
      - ./config/radarr:/config # Replace with the path to your config directory+
      - ./downloads:/downloads
      - ./movies:/movies
      - /mnt/merged:/mnt/merged # Replace with the path to your media files
    ports:
      - 7878:7878
    restart: unless-stopped

  sonarr:
    image: linuxserver/sonarr
    container_name: sonarr
    environment:
      - PUID=1000 # Replace with your user ID
      - PGID=1000 # Replace with your group ID
    volumes:
      - ./config/sonarr:/config # Replace with the path to your config directory
      - ./downloads:/downloads
      - ./tv:/tv
      - /mnt/merged:/mnt/merged # Replace with the path to your media files
    ports:
      - 8989:8989
    restart: unless-stopped

  emby:
    image: emby/embyserver:latest
    container_name: emby
    environment:
      - UID=1000 # Replace with your user ID
      - GID=1000 # Replace with your group ID
    volumes:
      - ./config:/config # Replace with the path to your Emby config directory
      - /mnt/merged:/mnt/media # Replace with the path to your media files
    ports:
      - 8097:8096
      - 8920:8920 # Optional for HTTPS
    restart: unless-stopped

networks:
  default:
    driver: bridge

disable ssh password login

Sun, 01 Sep 2024 00:00:00 +0000

It is widely recommended to disable password authentication unless you have a specific reason not to. Instead you should use SSH keys.

To disable password authentication, look for the following line in your sshd_config file at /etc/ssh/sshd_config

#PasswordAuthentication yes

replace it with a line that looks like this:

PasswordAuthentication no

Now check for /etc/ssh/sshd_config.d/50-cloud-init.conf

If the file exists make the same change in that file as well. You may go ahead and delete the file altogether as well but some users report that it gets recreated.

Once you have saved the file and restarted your SSH server, password authentication should be disabled.

kworker high cpu usage fix

Sun, 01 Sep 2024 00:00:00 +0000

On my Debian 12 machine, I’ve been having a issue that, kworker take up 80%-100% of a CPU non-stop. After hours of searching through the web for a solution, what fixed it was on a post on askubuntu from 2018. So it seems like it’s not a new thing. I’ll link the relevant discussion below as well for reference.¹

So what you do is disabling dynamic USB power management in kernel.

In order to do that you have to add usbcore.autosuspend=-1 to your kernel’s boot command line:

GRUB_CMDLINE_LINUX_DEFAULT="<existing stuff> usbcore.autosuspend=-1"</existing>

To do that permanently, open the grub configuration file

sudo nano /etc/default/grub

and add the flag below:

usbcore.autosuspend=-1

Next you have to update your grub via

sudo update-grub

After reboot, it should be fixed.

https://askubuntu.com/questions/1044872/ubuntu-16-04-kworker-using-high-cpu-constantly ↩︎