Hi, There is no options at all to 'shape' upload+download=fixed bandwidth without using IMQ virtual device ? So if somebody is planning to give customer fixed bandwidth regarless if a customer is 'uploading' or 'downloading',he (or she:) MUST use IMQ virtual device ? Anybody did some tests of configuring same traffic shaping with and without using IMQ device ? How IMQ device (kernel) will behave on heavy load traffic (like VoIP,gaming,other 'real-time' traffic) comparing on mashine without using IMQ virtual device ? Does anybody did some test on using u32 filter and fw filter ? How does each one behave on same mashine,same kernel,same traffic shaping configuration ? f2zubac On Mon, 24 Feb 2003, Stef Coene wrote: > On Monday 24 February 2003 21:19, hanumantha kavuluru wrote: > > Hi All, > > Thanks for the help. Stef, our product is already out in the market with > > 2.4.18 kernel. So if we change the kernel, we will have to go through a > > full QA cycle which is something we don't want to do :). > Ok > > > BTW, I would like > > to have global limits on the Bandwidth per user irrespective of which > > interface the traffic might go.Can i do this without using IMQ? > So you want up + down = fixed bandwidth? Then you need IMQ. > > > Also i want > > to do my own classification by setting the skb->priority field with the > > appropriate class ID. Is it possible to use this kind of classification > > with IMQ and HTB? > I don't know the internals of tc, but I think you can do. > > Stef > > -- > > stef.coene@xxxxxxxxx > "Using Linux as bandwidth manager" > http://www.docum.org/ > #lartc @ irc.oftc.net > > _______________________________________________ > LARTC mailing list / LARTC@xxxxxxxxxxxxxxx > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/ >
