[Click] delay

[Philippe De Neve]

Hi Joachim,

Tnx for your reply.

I don't think this will realize what I want because the delayshaper delays
every packet for a certain time. So if a burst of 10 consecutive packets
arrive in a queue before the delayshaper, every packet is delayed for a
certain time resulting in the burst being spread out more. So the delay
shaper not only delays the packets but also changes the shape of the stream.
What I want is something that delays but NOT changes the shape a stream.
I want to do this to test interactive applications to check how much delay
is allowed between end users.

regards, Philippe.



----- Original Message ----- 
From: "J Klein" <JKlein at gmx.com>
To: "'Philippe De Neve'" <Philippe.deneve at intec.ugent.be>;
<click at amsterdam.lcs.mit.edu>
Sent: Friday, January 07, 2005 6:28 PM
Subject: RE: [Click] delay


> Hi Philippe,
>
> Delayshaper (pull, pull) as well as delayunqueue (pull, push) use the
> timestamp of the packet as basis for the added delay if the timestamp is
> set.
> If your packet source does not already set the timestamp of the packet you
> can put a Settimestamp as a first element in your graph, and delayshaper
or
> delayunqueue should do the best they can to give you a constant delay
> regarding the arrival time of the packets - eg at the end of your click
> graph you put a Queue(...)->Delayshaper(...)->ToDevice(...). Isn't that
what
> you want or did I miss something?
>
> Regards
> -joachim
>
> -----Ursprüngliche Nachricht-----
> Von: Philippe De Neve [mailto:Philippe.deneve at intec.ugent.be]
> Gesendet: Freitag, 7. Januar 2005 14:28
> An: click at amsterdam.lcs.mit.edu
> Betreff: [Click] delay
>
> Hi,
>
> is there a possibility to construct, with standard elements, a 'distance'
> delay? I know about the delayshaper but this element delays every packet
> with a certain time resulting in a different delta between packets.
> My intention is to create a delay so the stream on the output of click is
> exactly the same as the input stream only DELAY seconds delayed.
> Has anybody done this?
>
> regards, Philippe.
>
>
>
>
>
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>