The end users report issues with datastore reachability between the newly installed virtual machines (VMs) and the storage array. The VMs are deployed on a Cisco C-Series server, directly connected wi

Fibre Channel over Ethernet (FCoE) traffic must not be carried on a native VLAN according to Cisco best practices and FCoE standards, as it can cause unexpected behavior, including traffic loss. By separating FCoE traffic to a dedicated non-native VLAN (e.g., VLAN 10) and using VLAN 99 for data traffic, FCoE frames will be properly tagged and processed, resolving the datastore reachability issue.

Configuring host-facing FCoE ports as spanning-tree port type edge speeds up port transitions but does not resolve the fundamental issue of FCoE traffic being carried on a native VLAN.

While FCoE VLANs must be associated with a VSAN, the core problem is the use of the native VLAN for FCoE traffic, not merely the configuration of the VSAN database.

Configuring FCoE VLAN traffic on a separate interface is often a good practice, but the immediate problem described is the use of the native VLAN on the same interface, so changing the VLAN ID on the existing FCoE interface is the direct solution.