I'm trying to achieve a load-balanced NIC team on a Windows Server 2016 machine. This is a ProLiant using a 4x 1Gb 311FLR interface card which as of yesterday has the latest driver installed (100.20.0.64). The NIC team has been created in Windows Server. All four NICs have been added to the team and the following settings have been applied:
- Teaming Mode - LACP
- Load Balancing Mode - Dynamic
The server has been connected to a HP 5406Rzl2 which has the following config for the ports:
- Mode - Auto
- Flow Control - Enabled
- LACP Status - Disabled
- Trunk Type - LACP
We also have a NetApp connected to the switch using a 4x 1Gb trunked connection with the same config as the server ports.
The Windows server is running Backup Exec, when I run a backup and monitor the stats for the ports on the switch, only one of the server NICs is being used and this is only 0.5%
The backup job is extremely slow, even when tweaking the settings in BE.
The NetApp, which is configured the same on the switch end, is showing utilisation of all four NICs as I would expect.
I know I'm not going to get the full 4Gb but I would expect a lot more than 0.5%
Reading through a description of the Server teaming settings (https://www.starwindsoftware.com/blog/windows-server-2016-nic-teaming-functionality) makes me think I have the correct configuration on the server end but clearly, something is not working as it should be.
So, my question is: How do I get load balancing to work at the server end?
0 Answers