This message was deleted.
# harvestera
adamant-kite-43734
03/08/2022, 5:05 AMThis message was deleted.
f
full-football-64901
03/08/2022, 9:30 AM@tall-match-44242
i assume your "management network" means the "harvester-mgmt" (a.k.a you chose the interface while installing)
If so, we always recommend to have two or more interface or harvester, separate "harvester-mgmt" and "VLAN network".
you could refer to the docs page, but you need to choose different interface for ClusterNetwork.
The best practice just show the "VLAN" inside and outside harvester, and how it works with external switch.
t
tall-match-44242
03/18/2022, 3:09 AMHi @full-football-64901, thanks for the information. We’re using Dell switches and the config looks a little different but I think it’s the same as your example. Yes, in Harvester I do have it setup as VLAN 1 and I’m able to deploy VMs that are on the same network as the Harvester host.
As for the VLAN network interface I might not be understanding something or not explaining it well, I’ll give it another try. When installing Harvester I selected network card 1 and network card 2 to create the “harvester-mgmt” interface on the host. Network card 1 goes to network switch A and network card 2 goes to network switch B. Now maintenance can be performed on switch A or B and the Harvester host will stay connected to network. The Harvester host does have network card 3 and 4, so I would like to create the same setup for the VM network.
The first attachment is a screenshot of the install for the “harvester-mgmt”. The second attachment is of the vlan configuration in settings. Is it possible to configure en03 and en04 with 802ad (LACP) to give VLANs network redundancy between network switch A and B?
Thank you for your time.
tall-match-44242
03/18/2022, 4:21 PMHi @full-football-64901, I’ll do my best to answer. Dell professional services set it up and I’ve made some basic changes, like adding a tagged vlans on server ports. I looked up MLAG and it looks similar to what we have but I don’t know if it’s the same. What we’re using has been referenced as VLT. Attached is screenshot from the internet that best illustrates our setup. The network is spine-leaf and each server is connected to a leaf switch using the LACP protocol.
Thanks for taking time to look at this, no rush using the harvester-mgmt network is sufficient for our POC. I’m trying to understand how best we could deploy Harvester into production in our environment.
Copy code
Here is an example of the port configuration on a leaf switch connecting to a server.
This is an example of VLT configuration
interface Port-channel 10
no switchport
portmode hybrid
switchport
lacp long-timeout
vlt-peer-lag port-channel 10
spanning-tree rstp edge-port
no shutdown
Here is an example of the 10G interface configuration to the server
interface Te 1/7
description <server_name>
port-channel-protocol lacp
port-channel 10 mode active
exit
no shutdown3 Views