Changes the implementation of Prometheus to use its Operator instead of the regular configmap implementation.
This allows the deployment of metrics through independent ServiceMonitors instead of a centralized ConfigMap.
This update is reflated in Grfana's datasources.
When working on the development cluster, setup a kustomize environment to enable or disable services dynamicly, and include some dev specific variables (LoadBalancerIP, domain name etc.)
The previously used file naming convention is no longer necessary as Kustomize already orders any request done to the fluster, without having to rely on the `--recursive` fly of the `apply` command.
When deploying anything on the cluster, a `kustomization.yaml` is to be used to allow for on-the-fly patches, and autonomous cluster configuration through native k8s file configuration. This commit adds the first iteration for such implementation, essentially fixing the issue of setting up the cluster from a clean base. Now everything is consolidated into one single file, which later applies desired resources. This also fixes having to copy and paste URLs into the console, as Kustomize allows one to put those URLs into the kustomization file.
A new traefik deployment has been added, and two ingressclasses have been set for the new instance, as well as the previous one. This allows the network to be split in two, one for external access and another for internal access. Each traefik deployment is connected to a loadbalancer requesting the IP necessary for each type of access.
Include networking management and configuration for compatibility with Prometheus, enable and make registry accessible, configure Prometheus accordingly
When loading Vaultwarden on the cluster, it has been judged to be better to have DB files locally, on top of the longhorn backup. This makes data persistency much easier on the long term.
