All of lore.kernel.org
 help / color / mirror / Atom feed
* [LARTC] ingress/egress shaping
@ 2003-01-27 16:45 Björn Snippe
  2003-01-27 17:20 ` Björn Snippe
  2003-01-27 17:57 ` Björn Snippe
  0 siblings, 2 replies; 3+ messages in thread
From: Björn Snippe @ 2003-01-27 16:45 UTC (permalink / raw)
  To: lartc

Hi,

one short question: my setup consists of
  [WWW]--[ISP's routerbox]--[Linux router/shaper]--[LAN]
Is it practical to create qdiscs/classes for both the inner and the 
outer NIC, in order to get complete control of both U/D traffic?
Or would I better use IMQ? (control only one of the NICs?)

Thanks,
Björn

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

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [LARTC] ingress/egress shaping
  2003-01-27 16:45 [LARTC] ingress/egress shaping Björn Snippe
@ 2003-01-27 17:20 ` Björn Snippe
  2003-01-27 17:57 ` Björn Snippe
  1 sibling, 0 replies; 3+ messages in thread
From: Björn Snippe @ 2003-01-27 17:20 UTC (permalink / raw)
  To: lartc

David Boreham wrote:
> With IMQ you'd have three NICs :)
> IMQ is a virtual NIC.
> So if you can achieve your goals with control on the two real NICs,
> do that.
Yes, that's clear to me (well, quite :)), but I read about the 
limitations of the ingress qdisc and that using a virtual imq device 
would give me the power to use classes even on incoming traffic.
I just wanted to be sure not to be doing something stupid <:)
So there's nothing wrong with my thinking?
What comes in from the WWW, I can shape on my inner NIC, and what leaves 
to the router/WWW, I can shape on the outer one?
What happens with all the traffic coming in on my inner NIC from my LAN 
that gets delayed/shaped in order to leave on my outer NIC?
Will my inner NIC drop it, or will my Linux-router collect more and more 
packets, swell up to giant size and explode? :)

Thanks for any answer,
Björn

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

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [LARTC] ingress/egress shaping
  2003-01-27 16:45 [LARTC] ingress/egress shaping Björn Snippe
  2003-01-27 17:20 ` Björn Snippe
@ 2003-01-27 17:57 ` Björn Snippe
  1 sibling, 0 replies; 3+ messages in thread
From: Björn Snippe @ 2003-01-27 17:57 UTC (permalink / raw)
  To: lartc

David Boreham wrote:
> Repeat: queueing is done on outbound interfaces.

Got it. :)
[...]
> I'd recommend that you setup a test machine with the tc
> stuff on it, and play around for a while. It's much easier
> to understand this stuff by performing experiments, using
> tcpdump , looking at the stats output from tc, and so on.

My 'test machine' is the up-and-running router for the dorm where I live 
in :)
It's running fine, and all the basic stuff works like I think it should, 
I just want to make sure I'm not wasting lots of performance by using a 
silly configuration...
Thanks a lot for reassuring me,

Björn

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

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2003-01-27 17:57 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-01-27 16:45 [LARTC] ingress/egress shaping Björn Snippe
2003-01-27 17:20 ` Björn Snippe
2003-01-27 17:57 ` Björn Snippe

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.