Series
Ordered reading paths through the posts. Seven of them, each starting at the beginning of a subject and ending where it stops being one.
Most posts here were written in the order the problem showed up, which is rarely the order you want to read them in. A series fixes that: same posts, put in the order that makes each one make sense.
Three of the seven start from nothing and are linked from the front door. The other four assume you already know what you came for.
Install first, then the Flux comparison, then app-of-apps, drift, progressive delivery, and getting the cluster back from Git.
Why you would run your own, what to buy, how to reach it from outside, what to run on it, and how to get it back when it dies.
Talos, NixOS against Talos for cluster nodes, and what a declarative machine is worth when an auditor asks.
GPG from the first key, then pass, then the YubiKey, then what happens to all of it when the maths changes.
From what a container actually is to running a cluster with no internet, with the failure modes that bite in between.
Why Linux on the machine you use all day, the defaults before the customisation, and the distro question last.
Metrics, logs and traces, then the dashboards and alerts someone actually reads, then the topology for more than one cluster.