Okay, authentik is up! Took a while, I was fighting against flux and the helm release because it deployed with the wrong StorageClass (I forgot to have that configuration ready before release.) Helm wasn't able to modify the PVC because they're immutable, updating the release has to wait for the initial release to succeed (which it won't) or timeout and flux is quiet on the reasons for all of this unless you know where to look lots of learning was had though!

Anyway, admin and personal user accounts created, MFA enabled. Got my first application integrated too! (actual budget)

What next? The world is my oyster... Probably gitea or semaphore. I'm hesitant to integrate services like jellyfin before I have more users onboarded and this gives me an opportunity to experiment with other edge cases like other providers and service accounts and such

Part of my project for adopting gitops in my homelab has been setting up git hosting. I've selected Gitea for this instead of Gogs because I thought I'd have an easier time bootstrapping fluxcd...

Now that I've had more time to sit with the docs and learn more about Gitea, I've noticed they have broad support to serve as package registries too! It seems like I can host helm charts and docker images alongside my infra as code! This is fantastic, it'll greatly simplify my architecture. Two birds, one container!

First time immediately jumping to #fluxcd installing a bunch of resources and charts without doing it manually first to get a good understanding of the process. Worked out amazingly well, really no need to even fiddle with manual chart installs and such anymore which is neat. #kubernetes #gitops

Anyone using terraform/opentofu for their homelab setups? Either on infra level or for CM?

I've made it a project for this year to get everything managed via gitops. I'm taking it step by step and as such I haven't locked down manual write access so that I can tinker with stuff and troubleshoot as needed.

I'm finding that I need a good way to spot state drift so that I get notified if I forget to correct things afterwards. I think this is going to be less frustrating than fighting against enforced state while I get my bearings.

I guess I could use a cron or a timer unit. Unless someone has any recommendations? I would like to manage all the TF using fluxcd eventually but I think it's too early to start enforcing desired state right now. I'm open to suggestions...?

Here as promised...

How to Setup Preview Environments with #FluxCD in #Kubernetes

https://developer-friendly.blog/blog/2025/03/10/how-to-setup-preview-environments-with-fluxcd-in-kubernetes/

This is a #gitops driven approach that leverages #k8s underneath.

What to expect from this blog post:

1. deploy PR-specific environments
2. establish unique URLs for each
3. auto-comment links on GitHub PRs
4. orchestrate it all with FluxCD CRDs

#sre
#devops
#blogging

Flux monorepo layout

Flux, or FluxCD, is a "set of continuous and progressive delivery solutions for Kubernetes that are open and extensible" and is my preferred way to do "GitOps" in my Kubernetes clusters. I like Flux because it handles Helm extremely well (I'm a big fan of Helm) and allows me to have a simple fallback if something goes wrong. In this post, I will go through the process of

https://dustinrue.com/2025/03/flux-monorepo-layout/

How do you feel like deploying preview environments to #kubernetes using #fluxcd?

A blog post is coming up...

Stay tuned

I've got a numerous ideas on what to write next on my #blog.

The top candidates include the following keywords:

#opentelemetry
#kubernetes
#golang
#fluxcd
#gitops

What I did in the last weeks (part 2):

All of my k3s clusters (and I have a few of them for $REASONS) are now running either #fluxCD or #argoCD. So everything #gitops now.

In general, I like the lightweightness of fluxCD, not having to run Redis and whatnot. But having a GUI is sometimes nice, even though the flux CLI is really easy to use and very intuitive.

And of course, #renovatebot is watching all repositories and sending merge requests to update things in the clusters. Nice!

I maintain the #fluxcd is the better solution (over #argocd) if you are heavily invested in helm. I prefer how it works with helm resources and I prefer that I can still manually take over a helm based resource extremely easily if I need to.

Learned about #KCL today at work, gives me data oriented golang-vibes.
I will probably move my #homelab to it, as it nicely integrates with #fluxcd. Should be easy and no more 'generic' helm charts!

Just a reminder that I will be in Stockholm two weeks from now for #DevFestStockholm25, to have a workshop on how to get up and running with #GitOps with #FluxCD. I will also be hanging around at the main conference the day after. Hope to see you all there

Has anyone gotten #c ilium installed via the #terraform helm provider and successfully handed responsibility over to #fluxcd (again with helm) post install? For some reason, when I do, it ignores my L2 settings. Manual, same manifests, all works great.

I'm super excited to introduce Flux Operator, a new component in the #FluxCD ecosystem, with a focus on high-level abstractions for #GitOps. https://control-plane.io/posts/flux-operator-introduction/

Officially being made redundant. Been a long time coming, but haven't been having much success in the current market.

If anybody is on the lookout for a dedicated and passionate Tech Lead / Full-lifecycle Software Developer with 12+ years experience would really appreciate a connection.

Dev - #Java #JVM #Spring #SpringFramework #Groovy #Gradle

Ops - #Kuberenetes #Istio #Helm #FluxCD #GitOps

Infra - #AWS #Terraform

#Linux #XP #FOSS

#GetFediHired
#JobSearch
#FediHire
#Fedijobs

Spontane Vortragsankündigung: Diesen Freitag (04.10.) hält @vespasian bei uns einen #Vortrag über das Thema "#GitOps in #Kubernetes mit #FluxCD". Los geht's um 19:30 bei uns im Space.

Mehr Informationen hier: https://www.hacknology.de/news/2024/09/28/vortrag-gitops/

Whoops. Accidentally made #fluxcd undeploy almost everything from my #K8S cluster by mistake

Tried to fix it, but Gitlab was unavailable .... because it runs on the cluster, so the gitops repo wasn't available.

I should fix that and put the repo outside the cluster

Soak in the Flagger knowledge from the 'Fearless Delivery with Flagger' ️ Enlightning stream last Thursday!

It is a lighboard collaboration between Stefan Prodan, the live audience, and myself! Go team! ♫

http://youtu.be/U6tG6n7XYFI

♫~We're going live NOW!

Want to reduce the risk of rolling out new software?

Join Stefan Prodan and me for ️ Enlightning and learn how Flagger helps you automate the release process for applications running on Kubernetes!

youtu.be/U6tG6n7XYFI

(⁠人⁠*⁠´⁠∀⁠｀⁠)⁠｡⁠*ﾟ⁠+ And maybe, if we're lucky, Stefan's dog Haru will make an appearance ｡⁠*ﾟ⁠+

Join ️ Enlightning and learn Flagger, a tool that automates the release process for apps running on Kubernetes!

Want to reduce the risk of introducing a new software version? Let's learn from Stefan Prodan this Thursday, Aug 15!

Click 'notify me':
youtu.be/U6tG6n7XYFI