All of lore.kernel.org
 help / color / mirror / Atom feed
From: Wang Jian <lark@linux.net.cn>
To: lartc@vger.kernel.org
Subject: Re: [LARTC] HTB + speed a bit lower than what shaped to
Date: Fri, 08 Apr 2005 11:04:42 +0000	[thread overview]
Message-ID: <20050408185820.02DD.LARK@linux.net.cn> (raw)
In-Reply-To: <42566135.4090200@megaserve.net>

Hi Jandre Olivier,

1. although the rate is 32Kbit, the ceil is also 32Kbit; you will get
very close 32Kbit, but no more that 32Kbit.

2. The kernel calculates rate based on the packet size, not payload
length. But your client, calculates rate based on the payload length.

If the traffic pattern is mainly small packet size, you will notice a
much lower rate.

On Fri, 08 Apr 2005 12:47:17 +0200, Jandre Olivier <jandre@megaserve.net> wrote:

> Hi Guys,
> 
> I have a little problem i dont understand why its happening,
> Initially i thought it was a network problem, but then i made a 
> crossover between me and the server and shaped on that interface with as 
> minimal as possible config
> 
> this my test:
> $TC qdisc add dev $INTERFACE root handle 1: htb default 30
> $TC class add dev $INTERFACE parent 1: classid 1:1 htb rate 10Mbit
> 
> # POOL1
> $TC class add dev $INTERFACE parent 1:1 classid 1:2 htb rate 512Kbit 
> ceil 512Kbit prio 0
> 
> # DEFAULT
> $TC class add dev $INTERFACE parent 1:1 classid 1:30 htb rate 10Mbit 
> ceil 10Mbit prio 3
> 
> # MAILSERVER
> $TC class add dev $INTERFACE parent 1:1 classid 1:12 htb rate 32Kbit 
> ceil 32Kbit prio 0
> 
> 
> $TC qdisc add dev $INTERFACE parent 1:30 handle 30: sfq perturb 10
> $TC qdisc add dev $INTERFACE parent 1:12 handle 12: sfq perturb 10
> 
> # dev eth1 - creating filters with marked packets
> 
> $TC filter add dev $INTERFACE parent 1:0 protocol ip handle 12 fw 
> classid 1:12
> $TC filter add dev $INTERFACE parent 1:0 protocol ip handle 30 fw 
> classid 1:30
> 
> iptables -t mangle -A POSTROUTING -d 192.168.2.1/32 -j MARK --set-mark 12
> iptables -t mangle -A POSTROUTING -d 192.168.2.1/32 -p tcp --sport 22 -j 
> MARK --set-mark 30
> 
> This all goes into the right classes, but when i do a download from the 
> mailserver, im suppose to get 4Kb, and it starts at 5Kb , which is 
> burst, and then move down towards 4kb then 3kb and stays around there, 
> but never goes to 4 on the dot.
> 
> Is this normal or should it be perfectly 4Kb/s ?
> 
> Am i doing something wrong?
> 
> Thanks
> 
> Jandre
> 
> 
> 
> -- 
> Regards
>  Jandre
> 
> 		"Some people are alive only because
> 		 it is illegal to kill them."
> 
> _____________________________________________________
> 
> 
> 
> _______________________________________________
> LARTC mailing list
> LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc



-- 
  lark

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

      reply	other threads:[~2005-04-08 11:04 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-04-08 10:47 [LARTC] HTB + speed a bit lower than what shaped to Jandre Olivier
2005-04-08 11:04 ` Wang Jian [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20050408185820.02DD.LARK@linux.net.cn \
    --to=lark@linux.net.cn \
    --cc=lartc@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.