Hi Tim I am doing the VCF training and am wondering why the stack is always being done for the transport? As you describe in the Understanding NSX architecture subject. Would it not be more efficent if the two Nics on that same subnet can directly communicate as opposed to going all down the stack, encapsulating each time down and the reverse on the way up?
Hi Ian,
Thanks for the comment. It seems you may be referring to the overlay network when asking about using the entire stack. The overlay subnets are virtual switches. The VM’s nics must communicate using the Tunnel to utilize the physical transport in order to traverse the two host ESX machines. “There is no spoon” connecting the logical switch between the two hosts. I hope this makes the encapsulation more clear.
Hi Tim I am doing the VCF training and am wondering why the stack is always being done for the transport? As you describe in the Understanding NSX architecture subject.
Would it not be more efficent if the two Nics on that same subnet can directly communicate as opposed to going all down the stack, encapsulating each time down and the reverse on the way up?
Regards Ian Bennett
ian@clearedps.com.au
LikeLike
Hi Ian,
Thanks for the comment. It seems you may be referring to the overlay network when asking about using the entire stack. The overlay subnets are virtual switches. The VM’s nics must communicate using the Tunnel to utilize the physical transport in order to traverse the two host ESX machines. “There is no spoon” connecting the logical switch between the two hosts. I hope this makes the encapsulation more clear.
LikeLike