Your server have two physical NICs and below is a description of your original networking configuration:
1. Disable one of the physical interface
2. Bind the second physical interface to Hyper-V, so that the interface is now actually only a virtual switch (a vitual interface is created as a result of this binding)
3. Configured the host to use virtual interface for external connection (this design will force host to use the same virtual switch that every other VM uses for external connections)
- VM's synthetic network adaptor is configured with static IP
- Saved the VM
- Backup the VM using WSB
- Restored the VM
- Started VM after restoration, VM's synthetic network adaptor was in DHCP state
- The problem symptom persisted even if we tried to rebooted the VM and disabled/enabled the problbematic network adaptor.
We use the following method to workaround the issue:
- Re-configured NIC to use previous assigned static IP
- We will get a warning message saying that this IP is already being assigned to a NIC. Do you want to want to use a different IP?
- Click No and close the TCP/IP properties
- NIC's connection will suddenly resume and all previous TCP/IP configuration returned to previous static state
The following design is rather recommended:
1. Configure the first physical interface for host's external connection
2. Bind the second physical interface with Hyper-V, so that the interface is now actually a virtual switch
3. Disable the virtual adaptor that was generated as a result of step 2 (this design separates the host traffic from all other VMs, allowing complete traffic isolation between the parent/child partition)
After making the above configuration changes, the problem symptom no longer happens.