In this paper, we consider a set of HTTP flows using TCP over acommon drop-tail link to download files. After each download, a flow waits fora random think time before requesting the download of another file, whose sizeis also random. When a flow is active its throughput is increasing with timeaccording to the additive increase rule, but if it su®ers losses created when thetotal transmission rate of the flows exceeds the link rate, its transmission rateis decreased. The throughput obtained by a °ow, and the consecutive time todownload one file are then given as the consequence of the interaction of allthe flows through their total transmission rate and the link's behavior.   We study the mean-field model obtained by letting the number of flowsgo to infinity. This mean-field limit may have two stable regimes: one with-out congestion in the link, in which the density of transmission rate can beexplicitly described, the other one with periodic congestion epochs, where theinter-congestion time can be characterized as the solution of a fixed point equation, that we compute numerically, leading to a density of transmission rategiven by as the solution of a Fredholm equation. It is shown that for certainvalues of the parameters (more precisely when the link capacity per user is notsignificantly larger than the load per user), each of these two stable regimescan be reached depending on the initial condition. This phenomenon can beseen as an analogue of turbulence in fluid dynamics: for some initial conditions,the transfers progress in a fluid and interaction-less way; for others, the connections interact and slow down because of the resulting fluctuations, which inturn perpetuates interaction forever, in spite of the fact that the load per useris less than the capacity per user. We prove that this phenomenon is presentin the Tahoe case and both the numerical method that we develop and simulations suggest that it is also be present in the Reno case. It translates intoa bi-stability phenomenon for the finite population model within this range ofparameters.