Question d’entretien chez Alacriti

When two HTTP requests are processed simultaneously and the second request completes before the first, how does the server make sure the responses don’t get interchanged?