I had a call yesterday with a customer who was asking me if he has a LB that support 100,000 connections how many inbound requests can handle.
I explained that the number of requests could be equal or greater since the established connections are 100,000. But he was asking if there was a way to correlate the “connections” with the “requests”
He has been working with an HTTP proxy ( nginx ) and I guess it accepts”request” not connections.
Also he was asking if the time that the request go from LB to the end back server is about 10ms ( which is sounds correct ) the round trip is about 50 ms….
If u have any info on any of these two items plz let me know.