I have my shared data on Longhorn, so for services that’s just longhorn as a PVC on rke2(k8s) and for clients I expose the NFS for mounts from a longhorn PVC to them to mount to.

How is the kubernetes (k3s/rke2) migration coming along?

Why not cloudinit instead of ssh?

I guess the project was abandoned but supposdly spirtually succeed by https://github.com/matomo-org/?footer

Great now i can Kompose convert my converted helm charts!

RKE2 is the next step up from k3s Same group mantains it (Rancher) but its built for bigger productuon uses (i.e. it deploys etcd instead of sqllite by default).

Reverse proxies! They can redirect based on the dns name used to get to them. This is based on layer 7 data though so just http(s) services and not multiple ssh tunnels for example.

k3s/rke2 (k8s distros) do it automatically with Traefik when you use the gateway or ingres apis

Also for DNS a fun option is sslip.io which lets you do <some service>-192-168-1-10.sslip.io and it redirects to your ip but with a dns name added.

Though your router likely has an easy way to add local entries for dns and also upstream for the rest (i.e. 8.8.8.8)