A rate-limiting middleware that ensures only the desired number of requests are being worked on concurrently.
