Post Archive

Hi all, As someone who’s spent countless hours setting up and managing infrastructure, I know how tedious and error-prone it can be to create a kubernetes clusters on different environments or service providers. But what if I told you there’s a way to make this process a whole lot easier? Enter Terraform - a powerful tool for automating infrastructure deployment - which many of you readers may already be aware of. So let’s skip the broad introduction and get right into the details.

Hi all, I wanted to move all my artifacts back into my homelab to be able to run it airgapped. To reduce the overhead of running multiple services to achieve this goal I’ve comitted myself on running the sonatype nexus repository manager. Through their broad community support multiple package types are supported by one solution. To survive the administration configuration hell of the repository manager and store a configuration as code within my git, I opted towards writing multiple terraform modules to configure my nexus. The following blog post shall give you a broad overview how I approached the issue and explain how to use my modules for this specific usecase.

Hi all, in my current role as platform engineer for the Hamburg Port Authority we need to keep the velocity high to provide a good service for our internal clients. As result most of our infrastructure is created through templates in any way.

Hi all, some projects and clusters may enter a maintenance mode in their lifetime and dont receive any updates, changes or even patches for a long time. If something like this happens it may be neccessary to rotate the certificates used by control planes. The control planes of vmware tanzu provide this functionality via kubeadm.

Hi all, with the version 0.13.2 of metallb comes a change in regard to layer2 ip announcements. Therefor it is now required to switch from the old configmap setup to a custom resource setup. To document my upgrade steps - here is a small write-up of things required to get metallb running again.

Hi all, as some of you may know, I’m interested in homelabbing and are hosting my own kubernetes cluster at home. As part of a good homelab it is essential to keep track of logs and metrics. The number one goto application for this usecase is often the kube Prometheus Stack, which is in my humble opinion a bit to big for my homelab in regard to memory, compute and storage footprint. While looking for alternatives I stumbled upon Victoria Metrics which seems to be a perfect fit for my usecase.

Hi all, I’m currently working on refactoring the way to set up kubernetes clusters within the infrastructure of my current employer. (Role: Platform Engineer) Due to growing configuration requirements and time-consuming decisions we’ve decided within our team that it is time to refactor the stack and try out something new. The current setup is based on flux-cd with a self-written templating software to render manifests based on a single configuration file. This configuration file is called config.yaml, who would have guessed that, and contains all critical information to bootstrap and deploy a new cluster environment. Basic manifests are provided from an internal kubernetes service catalog which is version pinned for a cluster. The rendered manifests are stored within a dedicated kubernetes-clusters repository (${cluster.name}/cluster.generated/${service.name}) and are initially deployed with a ci/cd approach to apply the tanzu kubernetes cluster and kickstart flux-cd on it. After the initial setup: flux-cd picks up the stored manifest files within the kubernetes cluster repository and installs everything.

Hi all, it’s been some time since I wrote my last article here. I switched jobs, started reading a lot more, worked on different projects so that blogging came way to short. With the new year I want to try to write at least a monthly entry with one new thing I learned and want to share. The following post describes the helm chart capabilities of kustomize and how to use it in your workflow.

Hi all, due to the awesome progress with proton and the integration in steam.. lets be honest… there is no need for Windows if you are not playing games that are competetiv nor secured via kernel-level anticheat. I’m also working most of my time with containers and kubernetes environments. Integrations of those two in windows are more or less not existing. Due to the wsl it gets better but it’s not quite native yet. Switching to an open operating system was the next logical step in my mind sooo.. here we are.

Hi all, last week we had a discussion on our discord server to play a minecraft modpack as group again and pretty much everyone knows - “this gets difficult to find one”. Due to the fact that in the last few years everyone hosted a modpack for the whole discord there are multiple playthroughs done in many modpacks for minecraft. After a kinda long search we settled on Manufactio by Golrith. Link to curseforge. From the looks of it - it is a pretty solid modpack but without any easy support for setting up a server.

Hi all,

Hi there!

When working with the Atlassian Bamboo CI Server it gets pretty quick annoying to setup a nice and readable build pipeline. Within my current project we’re having multiple consecutive maven tasks that build, test, deploy parts of the application.

Hi there!

Hey there, my professor keeps all its data in a seperate webstorage. The easiest way to get the files is to view files over the webbrowser at a specific domain. As a lazy person who dont wants to download all lecture files its a way easier to download it via wget in a resursive usage. For not downloading the parent folders you need to ignore them just with a secound parameter. The whole webstorage is secured with an simple httpauth

Etherpad lite is easy to extend via npm install ep. While extending etherpad you need to keep in mind that there are some plugins which create new public sites that are crawable by google, bing and co. As example a beloved plugin by me (https://github.com/JohnMcLear/eplist_pads). It creates nice lists of all your pads, but it creates public searchable ids under /list and /public. To fix that is pretty easy. You need just to edit your robots.txt file under /static/robots.txt. You can find it under etherpad-light/src/static. Just add these two lines at the bottom of the file.

I am currently active on developing and extending the openerp point of sale extension. Didn’t get anything of mine extensions into it.. The problem? Openerp initialises all Javascript-Code after rendering the whole page and its running the user-generated code befor running its own code. I found a very small and smart way to work around this problem with a recursive timeout caller. The script is nice as easy to understand. :)

Hey everyone, I am currently setting up my cubietruck to get my personal cloud running with owncloud and services like firefox sync to backup my browsing history. I ran into some stupid things caused by ubuntu.com behaviour with hosting sources for old distributions on ports.ubunut.com. Linaro for Cubietruck runs from start with Quantal Quetzal Ubuntu, what is fine. After setting up the system , my usual behavior is, to get the lasted updates and fixes but nothing happend just 404 errors from ubuntu.com. After checking things like connection error on IPv6 and other stupid things, I check back if there is still something up for quantal. Nope, nothing in there… What now? You just need to update your /etc/apt/source.list - I uploaded my new one to 2 hosts. phcn.de (http://paste.phcn.de/?i=1409570170) w8l.org (http://paste.w8l.org/kt82xg4erqo9)

Hello everyone, I am currently in my final learning phase for exams and stumbled about the topic, writing pseudocode for simple algorithms like selectsort,insertsort or bubblesort. I though about an hour of all three to get a nice and clean version done. I know there are “ready to use”-stuff on Wikipedia or other bulletin -boards… but to get an own version is somehow cool :)