FCM vNIC networking questions

Question

Hi all

I have done a fair bit of reading on multiple forums (Altaro/Naviko etc.) about hyper-v/FCM recommended network setup and I was hoping for a bit of clarification on a few points.

The recommendation seems to be to use a team and then create vlans/subnets per function eg. Live Migration, Heartbeat, Management so they each have their own vNIC on the same virtual switch.

How exactly does using vlans on a team ensure that one type of traffic eg Live migration does not starve other types of traffic eg Heartbeat specifically. Obviously using dedicated NICs per function would ensure that but would also be wasteful but what is the difference between having all functions in one
vlan vs a vlan per function ?

My understanding was that it is when you build the cluster that it picks up all available subnets and adds them in so I assumed if I wanted to add another vNIC to the hypervisors that for FCM to see it I would need to recreate the cluster.

But in the test lab I have I added a heartbeat vNIC to each hypervisor and FCM saw it fine. It has no adverse effect on the current setup.

Does this sound right or should the cluster be rebuilt for it to properly work ?

Many thanks for any insights.

Answer 1

VLANs would not ensure that one type of traffic does not starve another type. As you note, VLANs provide for isolation of traffic. If you want to ensure against starvation, you need to implement Quality of Service on either subnet or VLAN.

https://learn.microsoft.com/en-us/windows-server/networking/technologies/qos/qos-policy-top