TCP Window Shaping

[Thomas Liske]

Hi,

wlanmac schrieb:
> I wanted to start a thread on TCP Window Shaping. It is something I have
> had in the back of my mind for awhile. It would be great, for instance,
> if chilli would adjust the TCP window when limiting bandwidth as the
> first stage (then only discarding the overflow). It would help bandwidth
> throttling without the need for as many TCP retransmissions. 
> 
> But, general shaping down the TCP window could help the congestion of a
> network (as noted even in the Wikipedia article). So, I'm going to put
> in a new option in chilli called 'tcpwin'. When set to zero, the
> default, nothing happens. When set to a value greater than zero, all TCP
> frames will have their Window rewritten. 

this could already be done with iptables. How does traffic shaping work
in chilli atm? Does it just hard drop any packets when the maximum
allowed rate of a client has been reached? I would be more attired
inimplementing something like RED for Tcp or ECN for congestion
avoidance. Is there any paper descriping how to calculate the effect of
window shaping on Tcp rates? Maybe there is already a good way to do
traffic shaping with iptables but I didn't check it, yet.


Regards,
Thomas L.

-- 
support at ibh.de                              Tel. +49 351 477 77 30
www.ibh.de                                  Fax  +49 351 477 77 39

-----------------------------------------------------------------------
Dipl.-Ing. Thomas Liske
Netzwerk- und System-Design


IBH IT-Service GmbH                         Amtsgericht Dresden
Gostritzer Str. 61-63                       HRB 13626
D-01217 Dresden                             GF: Prof. Dr. Thomas Horn
Germany                                     VAT DE182302907
-----------------------------------------------------------------------
Ihr Partner für: LAN, WAN IP-Quality, Security, VoIP, SAN, Backup, USV
-----------------------------------------------------------------------
       professioneller IT-Service - kompetent und zuverlässig
-----------------------------------------------------------------------