Re: ppp+vpn+htb

Linux Advanced Routing and Traffic Control

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi hiphin,


On Wed, 6 Apr 2005 13:39:40 +0200, "hiphin" <hiphin@xxxxxxxxxxxxx> wrote:

> Wang, hi, little of thinking and listen what I found logical
> 
> Q1: Is ppp interface, as only one connection on my real eth0 ?
> 
> If answare is YES, then simple SFQ on eth0 is good ?
> ...and main difference SFQ/ESFQ is that SFQ watching connections, and ESFQ watching host IP... 
> 

You didn't tell if you use pptp, but I think you use pptp.

ppp interface in this case is a peusdo network interface, under it there
is IP connection between client and server. So your control the ip
connection's rate, and then you can control the ppp interface.

So if you want fairness amongst these interfaces, you can either enforce
fairness on connection ( 1 client only has 1 ppp connection) or enforce
fairness on host (1 client may have multiple ppp connection but you take
the client as one)


> Q2: I have 2.4.xx, kernel ESFQ is for 2.6 kernel only ?
> 
> 
> I need to limit max bandwidth for all users and fair share, do not need to have guaranteed-minimum
> 
> mathematic is simple: N*user <= 256kbps
> 
> soo if I have one user on vpn, hi gets 256kbps
> if I have 2 users on vpn downloading, both is getting max 128kbps
> and soo on...
> 
> (this is too simple to be ...)


I am not familiar with ESFQ. In my test, SFQ doesn't provide good
fairness. ESFQ should have some improvement, I think.

Any good implemented fair queue can achieve your goal.


-- 
  lark

_______________________________________________
LARTC mailing list
LARTC@xxxxxxxxxxxxxxx
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc

[Index of Archives]     [LARTC Home Page]     [Netfilter]     [Netfilter Development]     [Network Development]     [Bugtraq]     [GCC Help]     [Yosemite News]     [Linux Kernel]     [Fedora Users]
  Powered by Linux