All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Sebastian 'spax' Pape" <pape@rbg.informatik.tu-darmstadt.de>
To: lartc@vger.kernel.org
Subject: [LARTC] Some problems with limiting downstream
Date: Wed, 13 Jun 2001 00:33:35 +0000	[thread overview]
Message-ID: <marc-lartc-99239793606978@msgid-missing> (raw)

hi!

I'm trying to control down- and upstream with a 2.4.3 kernel, iptables
and tc. I managed to limit the upstream and it works perfect, but
somehow my downstream traffic seems not to be affected from traffic
controlling. I'm not sure if I missunderstood something or perhaps
only mixed up some parameters - I tried several configurations, but
it's always the same: the traffic is filtered to the correct class,
but there's no limitation. The "strongest" connection can always
allocate allmost all downstream traffic.

So here's the setup - perhaps someone can give me a hint?

# DDEV="dev eth0"
# DBAND="bandwidth 768Kbit"
# DOPT="maxburst 20 avpkt 1000"

eth0 is the device to the lan, the average pakte size should be
arround 1000 (I got this from some netfilter byte counters) - but I'm
not sure about maxburst - I just took ist from the HOWTO - what's that
for?

# $TC qdisc add $DDEV root handle 1: cbq bandwidth 100Mbit avpkt 1000
mpu 64

#        $TC class add $DDEV parent 1:0 classid 1:1 cbq bandwidth
100Mbit rate 100Mbit allot 1492 weight 10Mbit prio 8 $DOPT
#        $TC class add $DDEV parent 1:1 classid 1:2 cbq bandwidth
100Mbit rate 768Kbit allot 1492 weight 77Kbit prio 6 $DOPT bounded
#        $TC class add $DDEV parent 1:2 classid 1:3 cbq $DBAND rate
688Kbit allot 1492 weight 70Kbit prio 1 $DOPT
#        $TC class add $DDEV parent 1:3 classid 1:100 cbq $DBAND rate
384Kbit allot 1492 weight 38Kbit prio 4 $DOPT
#        $TC class add $DDEV parent 1:3 classid 1:200 cbq $DBAND rate
120Kbit allot 1492 weight 12Kbit prio 2 $DOPT
#        $TC class add $DDEV parent 1:3 classid 1:300 cbq $DBAND rate
120Kbit allot 1492 weight 12Kbit prio 3 $DOPT
#        $TC class add $DDEV parent 1:3 classid 1:400 cbq $DBAND rate
64Kbit allot 1492 weight 12Kbit prio 5 $DOPT

#        $TC qdisc add $DDEV parent 1:100 prio
# the same with [1:200, 1:300 and 1:400]

Now `tc -s qdisc` shows that no pakets have been dropped, there's no
overlimit, but for example the 1:300-class got 60M while 1:200 and
1:100 only got 10M traffic per class.

My first thought was, that if I use the full bandwidth there won't be
anything to limit, because the classes are always borrowing bandwidth
to and from each other - so I changed the rate of class 1:3 to 688Kbit
to have 80Kbit "to share" but that also had no effect.

Now I've no more ideas what to change - any ideas?

thanks a lot for your time :)

		Sebastian

-- 
Sebastian 'spax' Pape           | Your mouse has moved. Windows NT must be
mailto: sebastian@p-a-p-e.de    | restarted for the change to take effect.
pgp: http://p-a-p-e.de/pgp.asc  | Reboot now?  [ OK ]  
         --- Do you want to know more? http://www.p-a-p-e.de/ ---


_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/2.4Routing/

                 reply	other threads:[~2001-06-13  0:33 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=marc-lartc-99239793606978@msgid-missing \
    --to=pape@rbg.informatik.tu-darmstadt.de \
    --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.