Linux Advanced Routing and Traffic Control list
 help / color / mirror / Atom feed
From: King Yung Tong <tong@cs.dal.ca>
To: lartc@vger.kernel.org
Subject: Re: [LARTC] Ethloop?
Date: Fri, 07 Jun 2002 20:26:39 +0000	[thread overview]
Message-ID: <marc-lartc-102348165620545@msgid-missing> (raw)
In-Reply-To: <marc-lartc-102340461224067@msgid-missing>

Thank you, I put 100kbps to both 1:11 and 1:10. 1:10 (prio 0) get almost
50kbps form 60kbps excess and 1:20 (prio 1) get 10kbps form excess.
Is it the expected result?
I gussess the all 60kbps (excess) should go to prio 0 or by proportional
to rate in each class. The rate on the original class is 1:3.
./tc class add dev lo parent 1:2 classid 1:10 htb rate 30kbps ceil 100kbps
prio 0
./tc class add dev lo parent 1:2 classid 1:11 htb rate 10kbps ceil 100kbps 
prio 1




0.5 23490 23490 0 0 7830 7830 0 0 46202 46202 0 0
1.0 29308 29308 0 0 9471 9471 0 0 57537 57537 0 0
1.5 30066 30066 0 0 10311 10311 0 0 58370 58370 0 0
2.0 30364 30364 0 0 10190 10190 0 0 59133 59133 0 0
2.5 28144 28144 0 0 10251 10251 0 0 61029 61029 0 0
3.0 30859 30859 0 0 7907 7907 0 0 59899 59899 0 0
3.5 30563 30563 0 0 8694 8694 0 0 60245 60245 0 0
4.0 29243 29243 0 0 9873 9873 0 0 58596 58596 0 0
4.5 83553 74535 30 0 78956 16722 205 0 13905 13905 $5.0 97047 89562 20 0
95957 10635 438 0 3300 3300 0 05.5 100414 88939 38 0 100155 13212 553 0
783 783 0 06.0 101170 92980 20 0 101108 9802 570 0 186 186 0 0
6.5 98639 89751 38 0 98624 13015 577 0 44 44 0 0
7.0 99815 93172 20 0 99812 9755 578 0 10 10 0 0
7.5 63232 65056 0 0 39204 23397 41 0 3 3 0 0
8.0 54142 54575 0 0 22722 18971 3 0 1 1 0 0
8.5 51258 51361 0 0 21080 20189 0 0 0 0 0 0
9.0 50674 50698 0 0 20406 20195 0 0 0 0 0 0
9.5 49723 49729 0 0 20402 20351 0 0 0 0 0 0
10.0 51020 51021 0 0 20529 20517 0 0 0 0 0 0



On Fri, 7 Jun 2002, Martin Devera wrote:

> ethloop output (as text)
> 
> On Fri, 7 Jun 2002, King Yung Tong wrote:
> 
> >
> >
> > Pat
> >
> > On Fri, 7 Jun 2002, Martin Devera wrote:
> >
> > > Ok. It seems you generate only 40kbps by ethloop script. Then you
> > > can't expect your lines to go higher !
> > > Generate 100kbps to 1:11 and 1:10 ...
> > > devik
> > >
> > > On Fri, 7 Jun 2002, King Yung Tong wrote:
> > >
> > > > Here is the version and info.
> > > >
> > > > Version htb3_2.4.17
> > > >
> > > > [root@samwise htb]# ./tc -s -d class show dev lo
> > > > class htb 1:11 parent 1:2 leaf 30: prio 1 quantum 1024 rate 80Kbit ceil
> > > > 800Kbit burst 1701b/8 mpu 0b cburst 2623b/8 mpu 0b level 0
> > > >  Sent 130048 bytes 127 pkts (dropped 0, overlimits 0)
> > > >  rate 254bps
> > > >  lended: 100 borrowed: 27 giants: 0
> > > >  tokens: -60449 ctokens: 12800
> > > >
> > > > class htb 1:1 root rate 800Kbit ceil 800Kbit burst 2623b/8 mpu 0b cburst
> > > > 2623b/8 mpu 0b level 7  Sent 730112 bytes 713 pkts (dropped 0, overlimits
> > > > 0)
> > > >  rate 1426bps 1pps
> > > >  lended: 81 borrowed: 0 giants: 0
> > > >  tokens: 12800 ctokens: 12800
> > > >
> > > > class htb 1:10 parent 1:2 leaf 20: prio 0 quantum 3072 rate 240Kbit ceil
> > > > 800Kbit burst 1906b/8 mpu 0b cburst 2623b/8 mpu 0b level 0
> > > >  Sent 360448 bytes 352 pkts (dropped 0, overlimits 0)
> > > >  rate 704bps
> > > >  lended: 298 borrowed: 54 giants: 0
> > > >  tokens: -36462 ctokens: 12800
> > > >
> > > > class htb 1:2 parent 1:1 rate 320Kbit ceil 800Kbit burst 2008b/8 mpu 0b
> > > > cburst 2623b/8 mpu 0b level 6
> > > >  Sent 490496 bytes 479 pkts (dropped 0, overlimits 0)
> > > >  rate 958bps
> > > >  lended: 0 borrowed: 81 giants: 0
> > > >  tokens: -62221 ctokens: 12800
> > > >
> > > > class htb 1:12 parent 1:1 leaf 40: prio 0 quantum 6144 rate 480Kbit ceil
> > > > 800Kbit burst 2213b/8 mpu 0b cburst 2623b/8 mpu 0b level 0
> > > >  Sent 239616 bytes 234 pkts (dropped 0, overlimits 0)
> > > >  rate 426bps
> > > >  lended: 234 borrowed: 0 giants: 0
> > > >  tokens: 15867 ctokens: 12800
> > > >
> > > >
> > > >
> > > > result for all line with 100kbps (pic5)
> > > >
> > > > 0.5 23947 23947 0 0 7982 7982 0 0 45228 45228 0 0
> > > > 1.0 29036 29036 0 0 9571 9571 0 0 57118 57118 0 0
> > > > 1.5 29318 29318 0 0 10107 10107 0 0 59489 59489 0 0
> > > > 2.0 30522 30522 0 0 10253 10253 0 0 59327 59327 0 0
> > > > 2.5 28197 28197 0 0 10280 10280 0 0 61159 61159 0 0
> > > > 3.0 30910 30910 0 0 7927 7927 0 0 59985 59985 0 0
> > > > 3.5 30991 30991 0 0 7935 7935 0 0 60665 60665 0 0
> > > > 4.0 28994 28994 0 0 9598 9598 0 0 59183 59183 0 0
> > > > 4.5 30150 30150 0 0 10034 10034 0 0 14044 14044 0 0
> > > > 5.0 28662 28662 0 0 10458 10458 0 0 3333 3333 0 0
> > > > 5.5 30876 30876 0 0 10506 10506 0 0 791 791 0 0
> > > > 6.0 28540 28540 0 0 10523 10523 0 0 188 188 0 0
> > > > 6.5 29107 29107 0 0 10154 10154 0 0 45 45 0 0
> > > > 7.0 30387 30387 0 0 8500 8500 0 0 11 11 0 0
> > > > 7.5 45085 45085 0 0 18230 18230 0 0 3 3 0 0
> > > > 8.0 49523 49523 0 0 19855 19855 0 0 1 1 0 0
> > > > 8.5 49546 49546 0 0 20405 20405 0 0 0 0 0 0
> > > > 9.0 50916 50916 0 0 18643 18643 0 0 0 0 0 0
> > > > 9.5 50410 50410 0 0 20082 20082 0 0 0 0 0 0
> > > > 10.0 48597 48597 0 0 20453 20453 0 0 0 0 0 0
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > On Fri, 7 Jun 2002, Martin Devera wrote:
> > > >
> > > > > What version of HTB do you use ? Send me
> > > > > tc -s -d class show dev xxx
> > > > > outoput after test pic5 generation.
> > > > > devik
> > > > >
> > > > > On Fri, 7 Jun 2002, King Yung Tong wrote:
> > > > >
> > > > > > Thank you again for you help!
> > > > > > Attach is the pic5 for the ceil equal to 100kbps for all line.
> > > > > > 	      pic10limit for the ceil equal to 100kbps except 1:20.
> > > > > >
> > > > > > My objective is to show if not prio is set the excess should allocate
> > > > > > according to rate. If the prio is set, the high prio should get more.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Fri, 7 Jun 2002, Martin Devera wrote:
> > > > > >
> > > > > > > > Thank you for you answer, 1:10 is incresed but 1:20 is also increased and
> > > > > > > > increase to the specific rate if I add 100kbps to every line.
> > > > > > >
> > > > > > > sounds weird. With higher prio 1:10 should get 90k and 1:11 10k.
> > > > > > >
> > > > > > > > In my case, I would like to put all the extra to 1:10 only, is that means
> > > > > > > > I have to give 1:11 ceil to 10kbps. If it is, is that means I don't need
> > > > > > > > prio paramter? How prio parameter works?
> > > > > > >
> > > > > > > your understanding is ok, tge ceil 10k is correct - the prio is
> > > > > > > meant for lowering delay and change in excess distribution is
> > > > > > > only "secondary" product.
> > > > > > >
> > > > > > > Port the resulting graph somewhere so I can look at it.
> > > > > > >
> > > > > > > devik
> > > > > > >
> > > > > > > _______________________________________________
> > > > > > > LARTC mailing list / LARTC@mailman.ds9a.nl
> > > > > > > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
> > > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > > >
> > >
> > > _______________________________________________
> > > LARTC mailing list / LARTC@mailman.ds9a.nl
> > > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
> > >
> >
> >
> >
> 
> _______________________________________________
> LARTC mailing list / LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
> 

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

  parent reply	other threads:[~2002-06-07 20:26 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-06-06 23:02 [LARTC] Ethloop? King Yung Tong
2002-06-07 18:02 ` Martin Devera
2002-06-07 18:36 ` King Yung Tong
2002-06-07 18:45 ` Martin Devera
2002-06-07 18:57 ` King Yung Tong
2002-06-07 19:06 ` Martin Devera
2002-06-07 19:47 ` Martin Devera
2002-06-07 19:47 ` King Yung Tong
2002-06-07 19:53 ` King Yung Tong
2002-06-07 20:03 ` Martin Devera
2002-06-07 20:17 ` King Yung Tong
2002-06-07 20:19 ` Martin Devera
2002-06-07 20:26 ` King Yung Tong [this message]
2002-06-07 20:48 ` Martin Devera
2002-06-07 20:53 ` King Yung Tong

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-102348165620545@msgid-missing \
    --to=tong@cs.dal.ca \
    --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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox