Hetzner and their FailoverIP. Who used?

Question

Hetzner

Hello.
We collect a Proxmox HA cluster of 3 nodes from Hetzner.
2 nodes - under the VM
1 - storage
You can ask other hosters that all servers be in the same VLAN, and then if one of the nodes under the VM falls, the VMs themselves, along with their IP addresses, will quietly move to a spare node and continue to work.
Hetzner does not have such a service, they give FailoverIP.
Who used this FailoverIP, please explain how it still works?
As I see it now:
suppose we have 30 VMs, we take the /27 grid of these FailoverIPs and connect it to the main node.
Then, through vzctl, we distribute these FailoverIP addresses to all our VMs.
And now, at one not perfect moment, the main node falls.
All VMs go to the spare (all three servers are in a separate LAN, because we ordered additional cards and a switch) and in order for the network to work, we
assign the /27 FailoverIP network to the spare node (now it doesn’t matter how, whether it’s by hand or a script ) and all 30 VMs are still online and available on the Internet.
Am I wrong / not quite right / not right at all?
In general, tell me who is in the subject.

Answer 1

1. Hetzner can be asked for anything, up to switching servers between themselves into a separate physical LAN. They have more than adequate support.
2. FailoverIP - a stupid mechanism for quickly switching external IP from one physical machine to another.
With the subnet, by the way, there will be nuances, hetzner simply does not give ipv4 addresses, we got a maximum of 2-3 ip per physical server, and each time with a hard justification for "why".

Answer 2

as far as I remember in the admin panel, you need to transfer the IP filer to another server and then you can use it