We are planning to deploy CloudLens technology in an ESX virtual server environment of a customer, with vTAPs and a virtual packetbroker (vPB). This also involves a virtual NAM probe (AMD).
Now we need to decide how to sent the traffic from the vPB to the vProbe, so I am interested in available options to establish the connection between the two. So I am open to any thought on this.
The idea is to have the vPB within the same ESXh host as the vProbe, but it may be that they end up on separate hosts. (The ESX environment uses virtual distributed switches).
Within an ESX host one could create a vSwitch (or perhaps vDS) exclusively for the two entities to communicate, which I tested in our Lab. But there may be other and better solutions.
Looking for a possibility to use GRE tunnel between vPB and vProbe, I came across a few related posts, and read the article "NAM Probe (AMD) on AWS - a DYI approach" (https://community.dynatrace.com/community/pages/vi...) which gives a recipe on how to use GRE on an AMD, which may be a solution for either separate and same ESX host deployments.
After contact with Ixia they gave the GRE option as possibility, which is more dynamic option if NPB and AMD (NAM Probe) are not in same ESX host.
Other option is creating a vSwitch, flooding it from NPB and have vAMD read from the switch.
I assume a vDS can also be used.