KVM Transport Node: pnic and transport zones

[jcaamanosuse]

Hello

I have some questions related KVM transport nodes.

The NSX-T documentation seems to indicate that the interfaces assigned to the uplink profile for a transport node should be physical interfaces. Can they be vlan or bond interfaces? This kb article seems to indicate that linux bond interfaces will not work:

https://kb.vmware.com/s/article/67835

If this is the case, you need to dedicate physical nics to NSX-T, probably two if you want bonding. If you run other networks, like storage and management, which probably should be isolated from one another also, and for storage you probably also want some bonding, the you are now talking about minimum 5 total pnics.

Given this, is there any other way that you can share the pnics associated to the uplink profile for other networks? You can probably run your own vlan networks on top of the pnic provided to NSX-T but I don't think that makes much sense as then those network are slaved to however NSX-T is managing that pnic.

The pre-configured N-VDS seems to be a possible way to achieve this, but is there any documentation on how an N-VDS should be pre-configured or what that even means in the context of a KVM transport node?

Regarding transport zones, I can only configure one N-VDS although the documentation seems to indicate at least two should be possible:

You can add a maximum of four N-VDS switches for each configuration: standard N-VDS created for VLAN transport zone, enhanced N-VDS created for VLAN transport zone, standard N-VDS created for overlay transport zone, enhanced N-VDS created for overlay transport zone.

This could apply only to ESXi or BM hosts (since enhanced N-VDS only applies to ESXi) but the documentation does not do a good job in specifiying what applies to what.

Thanks.

[daphnissov]

KVM is just like ESXi when using as transport nodes in that NSX-T requires access to dedicated, physical NICs. You would want two for failover purposes, and KVM only supports the teaming policy of Failover Order. You can optionally use a LAG (which is still dedicated pNICs), but on KVM with LACP you can only have one.

[jcaamanosuse]

Thanks for the answer daphnissov.

In the NSX-T reference design

VMware® NSX-T Reference Design

Figure "6-8 KVM Compute Rack Failover Teaming" depicts non overlay traffic (for example, a storage network) added to the KVM N-VDS. But gives no further details on how to do so. While I can imagine different ways to do it, I am left wondering if it is ok for a non NSX-T entity to fiddle with the configuration of the N-VDS, which is actually a NSX-T managed OVS instance, where the pnics exist in a NSX-T controlled bridge.