Oops. Missed the context here. I just corrected the word Microservices to microservices deployed in kuberetes. So the whole article is for the microservices deployed in kubernetes environment, where concurrency will be handled by replicasof the microservice pods.

The request B will be routed to the second pod so no need for it to wait until request A completes. Hope this clarifies and welcome your further thoughts on it.

Cheers