This might be a newbie project so please be patient till the end of the post. I'm developing a demo data centre and have commissioned a couple of machines to act as nodes. My DC is defined as follows:
2 Machines running Hyper-V
2 (centos7, centos8) and 3 (centos7, ubuntu, manjaro) VMs running in 1st and 2nd Hyper-V server respectively
Monolith apps running inside these VMs (chatbot, ecommerce site, etc.)
Each of these VMs has an external network switch which means that they get an IP within the network and can be discovered/connected to from any device in the same network
I have prometheus node exporter running in each of the VMs and hyper-v exporter in both of the hypervisors that submit VM and hypervisor level metrics respectively. I also have exporters for application level metrics as well. I have a small machine that hosts prometheus in in it's .yml file, I've defined the endpoints of these machines and am able to graph the metrics
I wanted to know if instead of manually setting up the node and VM IPs, prometheus would be able to discover all the resources and applications, flag them appropriately (as a VM or hypervisor) and collect metrics from them. It tried going through the official doc but couldn't find any relevant resource. There's an active discussion in this forum too but didn't yield anything fruitful
Any help would be greatly appreciated. Thanks in advance
