• I encountered persistent issues with my container image build pipeline. The pipeline is run by my local GitLab CI instance. After struggling for a considerable period, I decided to address the problem by transitioning from a build setup relying on “docker” to a new setup built on “podman/buildah”. This required a complete overhaul of the build process. Unfortunately, this change broke the deployment of my website using GitLab CI.

  • I ran a rather decent monitoring system for my servers at home based on prometheus. My monitoring service sends an alert, if some file systems start filling up. Unfortunately, in most cases, I don’t know, what file(s) cause this mess. I use ncdu to find the largest files in the alerted filesystem. This short article shows how to install and use this small helper.