Applies to: Windows Server 2019, Windows Server 2022, and Windows Server 2025.
NIC Teaming combines adapters for resiliency without requiring switch-side aggregation.
Prerequisites
- Administrative access and an approved change window.
- A verified backup and documented rollback plan.
- A pilot environment or low-risk target for validation.
Step-by-step procedure
- Confirm both adapters and drivers support teaming.
- Connect adapters to approved switch ports.
- Create a switch-independent team.
- Test link failure during a maintenance window.
Commands
New-NetLbfoTeam -Name "Production-Team" -TeamMembers "Ethernet1","Ethernet2" -TeamingMode SwitchIndependent -LoadBalancingAlgorithm Dynamic
Get-NetLbfoTeamVerification
The team stays operational when either individual test link is disconnected.
Rollback
Remove the team and restore the original IP configuration to the documented physical adapter.
Security notes
- Use least privilege and record the change.
- Review relevant event logs after implementation.
- Test before production rollout.
Use teaming only on a supported network design
Switch-independent teaming does not require the physical switch to negotiate an LACP team, but the adapters still need a compatible network and VLAN design. Confirm that each member reaches the intended upstream network and that the driver and Windows Server version support the selected mode.
On a virtual server, guest adapter teaming is not a substitute for redundancy in the physical platform. Check what the provider exposes and supports before creating a team. On a Hyper-V host, distinguish classic LBFO teaming from Switch Embedded Teaming and use the method supported for the virtual switch and server release.
Creating a team can interrupt the connection that currently manages the server. Keep console access, record interface addresses, routes and VLANs, and plan which settings will belong on the team interface. Do not run the sample over your only remote path without a way to recover.
Test failover during an approved window by exercising the intended loss of a member link, then confirm client traffic and management access. A healthy team status does not prove both upstream paths carry the right traffic. Restore the recorded configuration if the pilot reveals routing, VLAN or application problems.
Related resources
Canadian Windows VPS and RDS hosting: managed Windows VPS hosting. Refer to Microsoft Windows Server documentation.