I see countless Youtube videos online that when they refer to cloud-init with proxmox they assume that you will be creating a template and inside that template you are going to attach a configdrive2 disk with the dynamic cloud init data that proxmox provides because the native Linux implementation doesnt work.
https://pve.proxmox.com/pve-docs/api-viewer/index.html#/nodes/{node}/qemu/{vmid}/config
Within the API documentation it specifies when you are using the Linux kernel by default it will use the "nocloud" datasource which I'm pretty familiar with using bare metal deployments where we use DHCP to give the machine an address and based on that address I have a metadata server to dynamically feed the cloud-init files like user-data and meta-data to the machine with the config useful for that address.
Why does Proxmox default to nocloud datasource but their nocloud implementation for Linux kernel doesn't work?
https://pve.proxmox.com/pve-docs/api-viewer/index.html#/nodes/{node}/qemu/{vmid}/config
Within the API documentation it specifies when you are using the Linux kernel by default it will use the "nocloud" datasource which I'm pretty familiar with using bare metal deployments where we use DHCP to give the machine an address and based on that address I have a metadata server to dynamically feed the cloud-init files like user-data and meta-data to the machine with the config useful for that address.
Why does Proxmox default to nocloud datasource but their nocloud implementation for Linux kernel doesn't work?