Does anybody already have any practical experience in using CNAs in an FCoE environment configured with NIC teaming options?
I want to dual-connect a dual-port CNA to two Cisco Nexus 5000 Series switches both configured as FCFs and connected to two different FC fabrics.
How does an active-standby, active-active transmit load-balancing or host port-channel configuration on the CNA effect FCoE traffic?
The dual-port CNA should be seen as two Ethernet NICs and two FC HBAs by the Windows OS. With a NIC teaming configuration in active-standby mode I assume that one CNA port in the failover team is passive while that same CNA port can be used for forwarding FC traffic encapsulated in FCoE. This should be possible because the multipathing software running on the host where the CNA is installed should still see two possible paths to a target using the two FC HBAs presented to the host OS assuming the target is dual-connected to both FC fabrics.
Can anybody say if this is correct? How will this apply to a host port-channel configuration on the CNA when the Nexus 5000 Series Switches are running in vPC mode?
Another aspect I am interested in relates to VLAN configuration. From my perspective the FCoE frames sent through the CNA ports should always be tagged with the corresponding FCoE VLAN ID discovered through the FIP VLAN discovery phase regardless of which VLANs are configured on a teamed virtual adapter. Is this correct? Thank you in advance for any comments.
I have this problem too.