A simple question came across the virtual watercooler yesterday and the experts weighed in with some great advice.

Question:

What are the recommended NIC teaming options when using VC? Any conditions where one setting would be preferable over another?

Answer:

There are no “recommended” NIC teaming option for Virtual Connect but ther is extra flexibility in defining your uplink sets, which give you HA you can’t get in the rackmount world.  To summarize:

You can make an NFT team from NICs that are serviced by the same or different VC Ethernet modules.  One NIC or the other is active at any given moment, and stays active until a link loss is detected.  Note – Smartuplink should be set on VC networks servicing NFT teams, so that an uplink failure translates to a downlink failure to the NIC, triggering NFT to fail over to the other NIC.  If all NICs in an NFT team are serviced by the same module, that module can be a single point of failure.  If serviced by separate modules, then you have better HA.

You can make a TLB team from NICs that are serviced by the same or different VC Ethernet modules.  If all NICs in an TLB team are serviced by the same VC module, that module can be a single point of failure.  If serviced by separate modules, then you have better HA.

SLB is NOT supported with any VC module. SLB=port channeling and VC doesn’t support it. Only NFT and TLB are supported.