Notes from the call and subsequent debugging

Correct me if anything is wrong, these are some insights and commands I found useful from the session which may have been obvious to you - maybe these can be incorporated in the docs somehow.

• kustomizations in default namespace
• helmreleases mostly in stackspin namespace
• can kill things with kubectl and kubernetes will restart them as the processes are stateless

kubectl controls kubernetes
which starts flux
which manages kustomizations and helmreleases

helm:

kubectl get helmrepositories.source.toolkit.fluxcd.io -A
kubectl get helmcharts.source.toolkit.fluxcd.io -A

to kill flux:

kubectl get pods -A
kubectl delete -n flux-system pods NAMES

logs:

kubectl describe helmrelease -n stackspin nginx
kubectl describe helmrelease -n kube-system metallb
kubectl describe helmrelease -A | less +G

reconcile:

$ flux get helmrelease -A
NAMESPACE   	NAME                  	READY	MESSAGE                                                 	REVISION	SUSPENDED
kube-system 	local-path-provisioner	True 	Release reconciliation succeeded                        	0.0.21  	False
cert-manager	cert-manager          	True 	Release reconciliation succeeded                        	v1.7.1  	False
kube-system 	metallb               	False	HelmChart 'flux-system/kube-system-metallb' is not ready	2.6.2   	False
stackspin   	nginx                 	False	dependency 'kube-system/metallb' is not ready           	       	False
$ flux reconcile kustomization metallb # always hangs up with kustomizations
$ flux reconcile helmrelease -n kube-system metallb
$ flux reconcile helmrelease -n stackspin nginx

post logs:

<details><summary>nginx logs</summary>

```yaml
```
</details>