There are several settings in Linux kernel to limit the number of concurrent connections and the timing aspect of it:
// the maximum number of open connections per second
// the maximum number of connections that can remain unclosed incase of delay in processing
// the maximum number of seconds the connection can remain open waiting for processing. And if longer than that, it gets cancelled.
Usually the timeout is longer than 1 second. In that cases, the concurrent connections and connections per second will differ a lot.