Hi Lysa,
the issue you’re seeing is a known behavior with Switch Embedded Teaming when the load‑balancing algorithm is set to Dynamic. Dynamic mode can hash flows across multiple physical NICs, which is efficient but can cause out‑of‑order packets in certain workloads, especially with TCP streams that are sensitive to ordering. To eliminate the retransmissions, you should switch the algorithm to HyperVPort, which pins each VM’s traffic to a single physical NIC port.
The correct way to change this is through PowerShell. Run the following command on the Hyper‑V host:
Code
Set-VMSwitchTeam -LoadBalancingAlgorithm HyperVPort -Name <YourSwitchName>
Replace <YourSwitchName> with the name of your SET vSwitch. You can confirm the current setting with:
Code
Get-VMSwitchTeam -Name <YourSwitchName>
After applying the change, traffic from each VM will be consistently mapped to one NIC, avoiding out‑of‑order delivery. Keep in mind that HyperVPort sacrifices some throughput balancing compared to Dynamic, but it is the recommended algorithm for workloads where packet ordering is critical, such as database or latency‑sensitive applications.
I hope you've found something useful here. If it helps you get more insight into the issue, it's appreciated to accept the answer. Should you have more questions, feel free to leave a message. Have a nice day!
Domic Vo.