Hi All,

I'm using HTB to control the data flows in and out of my server; I've noticed recently that a few clients connecting through the server 'seem' to be getting much more bandwidth than they're allowed.

What I've found is that HTB is correctly dropping the packets of their request to limit their speed, BUT, their TCP sessions don't seem to be scaling back as I would have expected.

For example, say one has a limit of 1Mbps; I've noticed on the WAN link that the traffic coming to them in consistently (for many hours) 2Mbps or greater; any traffic over their 1Mbps limit is correctly dropped by HTB, but their session should be scaling back.

Has anyone else encountered this problem? If so, is there any way to get around this?

Thanks
Anubis

Since the traffic shaping solution you're describing is implemented on the server and not the client, is it possible they are instantiating more than one session with the same server (2+ buckets) or perhaps instantiating multiple sessions with different servers?

Hi Dijetlo,

Yes, that appears to be what's happening; not much I can do about that I suppose.

Thanks
Anubis