Thank y'all for the first day of #Rejekts2025 with great talks and inspiring conversations!
I am excited that I got a spot for the #LightningTalks.
Looking forward to present you #Kubenix a tool leveraging #NixOS modules to declare #K8s workloads fully declarative.
I will also show how its #Helm integration essentially bridges the #CloudNative and #Nix ecosystem effectively, while offering additionally type safety.
See you at 18:15 in the hall #TheNash!
This is relevant because it means you cannot put an RFC3339-formatted timestamp in a label.
I learn new and annoying things about Kubernetes every single day.
For example, you cannot put a colon in a label.
This morning's *absolute* WTF moment:
Pod network traffic 100% packet loss outbound UNTIL I `tcpdump` it on the node, then it starts working fine.
What’s an overlooked topic that many "Getting Into DevOps’" guides miss? In my opinion, it’s networking. Understanding the basics of networking can make a huge difference, when troubleshooting. Thoughts?
This is a really great writeup on building Kubernetes controllers. I've just skimmed it but already it's the best/most helpful guide I think I've seen..... Need to read it more in-depth soon.
https://ahmet.im/blog/controller-pitfalls/?ref=cloudnative.now
HT @Marcus and the @CloudNativeNow newsletter!
Pulling the trigger, soon 
#DevOps #k8s #learnk8s #cncf #100DaysOfCode #100DaysOfDevOps #kubernetes
I'm new to #k8s and am wondering how to handle templating large amounts of config files. I couldn't find anything super useful in my search so I have an #Ansible sidecar I run to generate the kustomizations and config files. My most recent Ansible change was 30 lines, it resulted in changing 5,000 lines of YAML which will further be fed to Kustomize.
There has to be a better way?
I've heard about Helm, Yoke, KRO, and using an operator pattern. My understanding of those options is:
* Helm - My Org recommends avoiding (I don't know why)
* KRO - Not stable, but FFS neither is Kustomize
* Yoke - Almost kinda operator pattern
* Operator Pattern - This feels like reinventing a fucking config manager (ala #Ansible, #Puppet, #Chef, #Saltstack) for every fucking project. What new hell this is.
I'm hoping I'm missing something because the only workable flow for this workload is:
1) Create ansible roles/playbook to generate the kustomization.yaml and resources
2) Generate those kustomizations, check them into git
3) Use Kustomize via GitOps to expand the YAML even more
4) Push a metric fuckton of YAML to production
I'm losing my mind over here.
Went ahead and installed #harbor helmchart for the upcoming #docker api rate limits. Works like a charm. #TalosLinux #k8s
The new k8s bug has a lame name: IngressNightmare. *sigh* Where's the clever word play?
Too many people relying on simple appending Nightmare to the name of the attack surface these days... might as well get an LLM to name them if we're just going to copy the last bug name all over again.
Anyway, you can read about it here:
https://www.wiz.io/blog/ingress-nginx-kubernetes-vulnerabilities
Aufgrund der #Sicherheitslücke(https://www.golem.de/news/auch-in-deutschland-tausende-kubernetes-cluster-lassen-sich-aus-der-ferne-kapern-2503-194691.html) im #nginx von #kubernetes...
gerade mal mein #MicroK8s-#Cluster mit Hilfe dieser Anleitung(https://github.com/canonical/microk8s/issues/4964#issuecomment-2752344329) upgedated
Me 2 years ago: #Kubernetes is software with biggest regeneration ability.
Me 6 months ago: Wow, Longhorn has even better regeneration ability!
Me now: What the hell, Patroni has the best regeneration ability I have ever seen!
Welcome to our brave new world of self-fixing things...
Spoke too soon. Damn context object has no cache initialized so can't run a reconcile. How in the heck am I supposed to write a controller that can't check all the custom resources it's managing? This seems basic?
#Kubernetes #monitoring Made Easy with VictoriaMetrics Cluster
Our technical #guide walks you through setting up a VictoriaMetrics cluster using Helm charts, collecting k8s metrics via service discovery, and visualizing your data effortlessly.
What you'll learn:
Deploying #VictoriaMetrics in Kubernetes with #Helm
Scraping #metrics from #k8s components
Storing & visualizing data in VictoriaMetrics #tsdb
https://docs.victoriametrics.com/guides/k8s-monitoring-via-vm-cluster/
> After receiving a discovery response, the Kubernetes API server automatically registers all available types for this API group. Although this isn't considered common practice, you can implement logic that dynamically registers the resource types you need in your Kubernetes cluster.
Woa, this could be huge for the CRD management problem with operators and Helm
I didn't even know the API aggregation layer existed
https://kubernetes.io/blog/2024/11/21/dynamic-kubernetes-api-server-for-cozystack/
Man this kube-builder thing is a LOT of machinery for something pretty simple I'm trying to make. Can't even get it to run a reconciliation loop on startup cause it's all buried. Will check the docs more thoroughly tomorrow.
Interesting piece on the details of actual scale and shifting from Helm to CueLang with ArgoCD. Includes the why's as well as lessons learned.
Just released https://registry.terraform.io/modules/bcbrookman/kubeadm-k8s/aws to provision #k8s clusters on #aws using #terraform. It's early, but I'm also planning on publishing some #kubestronaut practice labs that use it.
arc) 
