[6bone] ipv6 traffic generator

[Abdul Basit]

Hello

i checked on native links, i found it is
reasonably better than normal ipv4 though.

I was using iperf actually, but i was not able
to find any tool that can generate graphs
based on iperf logs. can you point out some?

I am using tcptrace to analyze tcpdumps for now
to generate xplot/graphs ( but the problem is it
can handle only tcp traffic). for generating
ipv6 tcp traffic i use sendip but the problem
with sendip is, it can't generate any application
level packet (say http).

Can you point me some ipv6 traffic generators
that supports extensive features of ipv6, like
mobile ipv6 binding update time , fast handovers etc?
and generate graphs based on that info?

NS2 (Network simulator ) doesn't support many features
of Mobile IPv4 even, MOBIWAN is hard to setup,
though i set MOBIWAN correctly but i am not able
now to get trgraph read ns2 trace files generated
by MOBIWAN?

Kame MIPv6 support isn't properly documented,
USAGI has mobile ipv6 experimental support also
LINA/LIN6 isn't well-documented ?

Where to look ? are those all standards just exists
in theory ? Don't we have some elegant approach ? instead
of all softwares scattered around to support bits
of things ?

I beleive we need to have a project started on this
( like developing a IPv6 simulator supporting every bit
of information specified in most recent rfc's).

- basit
graduate student
wichita state univ.

On Wed, 20 Nov 2002, Jeroen Massar wrote:

> Abdul Basit wrote:
>
> > hello,
> >
> > are there any decent ipv6 traffic generators ,
> > free one's ofcourse ?
>
> As you didn't mention the reason for generating traffic
> I assume that you want to test IPv6 throughput.
> A good tool for this is IPerf as found on:
> http://dast.nlanr.net/Projects/Iperf/
>
> Good thing about Iperf is that it's IPv4 and IPv6 capable
> and as such one can test differences in throughput between 4 and 6.
> One will notice that unfortunatly most of the time IPv6 is at loss
> due to the current 6bone mess. When this has been cleared out, and
> there are a couple of groups pondering about this I know from testing
> on native links that IPv6 will have a real benefit over IPv4 when
> checking throughput.
>
> Greets,
>  Jeroen
>
> _______________________________________________
> 6bone mailing list
> 6bone@mailman.isi.edu
> http://mailman.isi.edu/mailman/listinfo/6bone
>