During a data-center refresh you are asked to build two new Windows Server hosts that will share files with client machines. Each host has two 10 GbE adapters that will be patched to two different top-of-rack access switches. You are not permitted to make any changes on the switch side, and additional throughput is not a design goal-the requirement is simply to keep the server online if one adapter, cable, or switch fails. Which network configuration on the servers BEST satisfies the availability requirement?
Create a switch-independent NIC team with one adapter active and the other in standby.
Enable SPAN/port mirroring on the switch ports and leave the adapters unteamed.
Configure IEEE 802.3ad (LACP) dynamic link aggregation across both adapters.
Assign separate IP addresses to each adapter and use DNS round-robin for client access.
Creating a switch-independent NIC team in active-standby (active-passive) mode allows the server to present one logical adapter while monitoring both physical links. If the active link, its cable, or the attached switch fails, traffic automatically fails over to the standby adapter without any switch configuration.
IEEE 802.3ad (LACP) teaming is switch-dependent; every member link must terminate on the same switch or on a stack that is configured for the aggregation. Putting the two adapters on separate, unmanaged switches would break the bundle, so it would not provide the required failover.
SPAN/port mirroring simply copies traffic for analysis; it offers no redundancy or automatic switchover.
DNS round-robin can distribute client connections but does not withdraw an unreachable IP from rotation, so clients may still receive the address of a failed link and experience outages.
Therefore, switch-independent active-standby NIC teaming is the most appropriate technique for link and switch failover in this scenario.