All of lore.kernel.org
 help / color / mirror / Atom feed
* [LARTC] Re: measuring queueing delay using iproute2
@ 2002-02-27 16:57 Don Cohen
  2002-02-28  9:02 ` bert hubert
  2002-02-28 16:51 ` Don Cohen
  0 siblings, 2 replies; 3+ messages in thread
From: Don Cohen @ 2002-02-27 16:57 UTC (permalink / raw)
  To: lartc


How about this:
 use tcpdump to record the packets of that class
 match up the incoming and outgoing packets
 take the difference in tcp timestamps

 > From: Radhakrishnan Vijay <vijayr13@yahoo.com>
 > To: lartc@mailman.ds9a.nl
 > Subject: [LARTC] measuring queueing delay using iproute2
 > 
 > I am using the tc functionality provided by iproute2
 > to setup classes and queueing disciplines. I have a
 > requirement to determine the queueing delay of each
 > packet belonging to a particular class so that i can
 > measure the average queueing delay over a time
 > interval for a particular traffic class. I would be
 > grateful if somebody could give me pointers on how i
 > possibly can achieve this. I suspect that i would have
 > to modify the source code of iproute2 , but am not too
 > sure on what segment of the code i should be
 > modifying.
It won't help to modify iproute2.  It's not doing the forwarding. 
It only tells the kernel what you want the kernel to do.
_______________________________________________
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] Re: measuring queueing delay using iproute2
  2002-02-27 16:57 [LARTC] Re: measuring queueing delay using iproute2 Don Cohen
@ 2002-02-28  9:02 ` bert hubert
  2002-02-28 16:51 ` Don Cohen
  1 sibling, 0 replies; 3+ messages in thread
From: bert hubert @ 2002-02-28  9:02 UTC (permalink / raw)
  To: lartc

On Wed, Feb 27, 2002 at 08:57:11AM -0800, Don Cohen wrote:
> 
> How about this:
>  use tcpdump to record the packets of that class
>  match up the incoming and outgoing packets
>  take the difference in tcp timestamps

Won't work - tcpdump sits before the packet scheduler, and not after.

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
http://www.tk                              the dot in .tk
http://lartc.org           Linux Advanced Routing & Traffic Control HOWTO
_______________________________________________
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] Re: measuring queueing delay using iproute2
  2002-02-27 16:57 [LARTC] Re: measuring queueing delay using iproute2 Don Cohen
  2002-02-28  9:02 ` bert hubert
@ 2002-02-28 16:51 ` Don Cohen
  1 sibling, 0 replies; 3+ messages in thread
From: Don Cohen @ 2002-02-28 16:51 UTC (permalink / raw)
  To: lartc

bert hubert writes:
 > On Wed, Feb 27, 2002 at 08:57:11AM -0800, Don Cohen wrote:
 > > 
 > > How about this:
 > >  use tcpdump to record the packets of that class
 > >  match up the incoming and outgoing packets
 > >  take the difference in tcp timestamps
 > 
 > Won't work - tcpdump sits before the packet scheduler, and not after.

No way.  For instance, the packet scheduler can decide to drop the
packet.  In which case it better not show up in tcpdump!
Tcpdump is after dequeue, just before giving the packet to the device.
_______________________________________________
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:[~2002-02-28 16:51 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-02-27 16:57 [LARTC] Re: measuring queueing delay using iproute2 Don Cohen
2002-02-28  9:02 ` bert hubert
2002-02-28 16:51 ` Don Cohen

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.