All of lore.kernel.org
 help / color / mirror / Atom feed
* [LARTC] HTB questions
@ 2002-06-06 17:26 Mancinelli Giovanni
  2002-06-06 18:06 ` Martin Devera
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Mancinelli Giovanni @ 2002-06-06 17:26 UTC (permalink / raw)
  To: lartc

Hello all,

I have some doubts about HTB algorithm:
Is it possible to deny client connection when bandwidth is saturated ?

I mean bandwidth manager manages  1Mbit  and link is full, can I stop new
connection ? 
can I guarantee a fixed minimum bandwidth to all connection already
established?

Thanks in advance for the answer.

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

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [LARTC] HTB questions
  2002-06-06 17:26 [LARTC] HTB questions Mancinelli Giovanni
@ 2002-06-06 18:06 ` Martin Devera
  2002-06-07 14:41 ` Mancinelli Giovanni
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Martin Devera @ 2002-06-06 18:06 UTC (permalink / raw)
  To: lartc

No. It would require tcp-qos cooperation.

On Thu, 6 Jun 2002, Mancinelli Giovanni wrote:

> Hello all,
>
> I have some doubts about HTB algorithm:
> Is it possible to deny client connection when bandwidth is saturated ?
>
> I mean bandwidth manager manages  1Mbit  and link is full, can I stop new
> connection ?
> can I guarantee a fixed minimum bandwidth to all connection already
> established?
>
> Thanks in advance for the answer.
>
> Best Regards,
> Giovanni
> _______________________________________________
> 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/

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [LARTC] HTB questions
  2002-06-06 17:26 [LARTC] HTB questions Mancinelli Giovanni
  2002-06-06 18:06 ` Martin Devera
@ 2002-06-07 14:41 ` Mancinelli Giovanni
  2002-09-14  6:56 ` curt brune
  2002-09-14  8:49 ` Stef Coene
  3 siblings, 0 replies; 5+ messages in thread
From: Mancinelli Giovanni @ 2002-06-07 14:41 UTC (permalink / raw)
  To: lartc

Hello Devik,
Do you have any information about?
Is there a tool does it on Linux?

Thanks for the answer,
Giovanni


-----Messaggio originale-----
Da: Martin Devera [mailto:devik@cdi.cz]
Inviato: giovedì 6 giugno 2002 20.06
A: Mancinelli Giovanni
Cc: 'lartc@mailman.ds9a.nl'
Oggetto: Re: [LARTC] HTB questions


No. It would require tcp-qos cooperation.

On Thu, 6 Jun 2002, Mancinelli Giovanni wrote:

> Hello all,
>
> I have some doubts about HTB algorithm:
> Is it possible to deny client connection when bandwidth is saturated ?
>
> I mean bandwidth manager manages  1Mbit  and link is full, can I stop new
> connection ?
> can I guarantee a fixed minimum bandwidth to all connection already
> established?
>
> Thanks in advance for the answer.
>
> Best Regards,
> Giovanni
> _______________________________________________
> 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/

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [LARTC] HTB questions
  2002-06-06 17:26 [LARTC] HTB questions Mancinelli Giovanni
  2002-06-06 18:06 ` Martin Devera
  2002-06-07 14:41 ` Mancinelli Giovanni
@ 2002-09-14  6:56 ` curt brune
  2002-09-14  8:49 ` Stef Coene
  3 siblings, 0 replies; 5+ messages in thread
From: curt brune @ 2002-09-14  6:56 UTC (permalink / raw)
  To: lartc

Hello,

I have a number of questions re: HTB

* Traffic Simulator

I saw the traffic simulator, ethloop -- very clever.  I was wondering, however, how to
simulate traffic destined for specific ports ?  In the HBT HOWTO there are gnuplots of
traffic destined for different ports.  

* Low rate shaping

My maximum uplink is 384kbits/sec .  I plan to shape HTTP,SMTP and other protos to some
fraction of the maximum.  Are there any special steps when shaping to these low rates ?  I
saw most of the examples are in the 5 to 10 mbit/sec range, roughly 30 times higher than
what I plan to do.

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

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [LARTC] HTB questions
  2002-06-06 17:26 [LARTC] HTB questions Mancinelli Giovanni
                   ` (2 preceding siblings ...)
  2002-09-14  6:56 ` curt brune
@ 2002-09-14  8:49 ` Stef Coene
  3 siblings, 0 replies; 5+ messages in thread
From: Stef Coene @ 2002-09-14  8:49 UTC (permalink / raw)
  To: lartc

On Saturday 14 September 2002 08:56, curt brune wrote:
> Hello,
>
> I have a number of questions re: HTB
>
> * Traffic Simulator
>
> I saw the traffic simulator, ethloop -- very clever.  I was wondering,
> however, how to simulate traffic destined for specific ports ?  In the HBT
> HOWTO there are gnuplots of traffic destined for different ports.
I never used ethloop myself, but I think it uses flowid's.  So if you create a 
htb setup with classes and filters, you can use the u32 filter to match 
packets with different ports.  Ethloop will generate the traffic in the 
flowid's you configured it for.

Or take a look at the P option :

P Number. Priocode set by setsockopt(sock,SOL_SOCKET,SO_PRIORITY,...). It is 
very convenient way to test qdisc because majority of classfull qdisc will 
use classid stored in priority field. To say that this flow should go to 
class 3:4f use P 0x3004f. This way you need no filters.

> * Low rate shaping
>
> My maximum uplink is 384kbits/sec .  I plan to shape HTTP,SMTP and other
> protos to some fraction of the maximum.  Are there any special steps when
> shaping to these low rates ?  I saw most of the examples are in the 5 to 10
> mbit/sec range, roughly 30 times higher than what I plan to do.
If you have class with very different rates, it's possible you can get 
warnings about quantum.  You can ignore them, but it's better fix it.  To be 
short, quantum = rate / r2q with r2q = 10 by default.  
And MTU (1500) < quantum < 60.000.
Quantum is the number of bytes a class my send IF 2 or more classes are 
fighting for bandwidth from the same parent.  You can overrule r2q when you 
add the htb qdisc and/or you can overrule quantum when you add a htb class.

Stef

-- 

stef.coene@docum.org
 "Using Linux as bandwidth manager"
     http://www.docum.org/
     #lartc @ irc.oftc.net

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

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2002-09-14  8:49 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-06-06 17:26 [LARTC] HTB questions Mancinelli Giovanni
2002-06-06 18:06 ` Martin Devera
2002-06-07 14:41 ` Mancinelli Giovanni
2002-09-14  6:56 ` curt brune
2002-09-14  8:49 ` Stef Coene

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.