I have a physical server (IBM HS22) with 2 NICs. I freshly installed Windows Server 2016 with AD Directory Services, DNS, DFS, and DHCP server roles. It's up-to-date with Windows Updates.
The two NICs are teamed using lbfoadmin as switch independent, dynamic, and no standby.
Problem
When the server reboots, many unexpected event log entries are recorded. All the errors are related to networking problems, or lack of networking.
Examples: DNS server could not bind to IP address, ADWS cannot service directory, DFS Rootxxx failed during initialization, ntpclient cannot find peer, Intersite Messaging failed, etc.
I receive 20-40 errors on each boot.
Notes
- If I break the team and just use one NIC (it doesn't matter which one), these errors do not occur
- If the team contains only one NIC, and I disable the other NIC, the problem still exists
- Changing other team settings such as standby adapter and type doesn't help
- This problem does not appear on similar hardware using 2012 R2
- Side note: The services seem to auto-recover because all the server roles seem to be working
I suspect the teaming layer is adding a delay to networking availability and is causing the services to fail because the network is not ready.
I really want the NICs teamed for fault-tolerance. Any ideas on how I can fix/workaround this problem?