[184349] in North American Network Operators' Group
Re: wanted: tool for traffic generation / characteristics / monitoring
daemon@ATHENA.MIT.EDU (Antonio Ojea Garcia)
Fri Oct 2 10:28:36 2015
X-Original-To: nanog@nanog.org
In-Reply-To: <20151001201119.GA27584@Mail.DDoS-Mitigator.net>
From: Antonio Ojea Garcia <aojea@hotmail.com>
Date: Fri, 2 Oct 2015 13:40:25 +0200
To: alvin nanog <nanogml@mail.ddos-mitigator.net>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
I guess you are looking for something like this
http://traffic.comics.unina.it/software/ITG/
D-ITG (Distributed Internet Traffic Generator) is a platform capable to
> produce traffic at packet level accurately replicating appropriate
> stochastic processes for both IDT (*Inter Departure Time*) and PS (*Packet
> Size*) random variables (*exponential, uniform, cauchy, normal, pareto,
> ...*).
>
>
2015-10-01 22:11 GMT+02:00 alvin nanog <nanogml@mail.ddos-mitigator.net>:
>
> hi matthias
>
> On 10/01/15 at 03:41pm, Matthias Flittner wrote:
> > Dear colleagues,
> >
> > Currently we are looking for a magic tool with which it is possible to
> > generate specific (realistic) traffic patterns between client and server
> > to analyze (monitor) traffic characteristics (jitter, delay, inter
> > arrival times, etc.).
>
> generating traffic and monitoring traffic is usually not done
> by the same apps .... there's hundreds of monitoring apps
> and hundreds of traffic generators
>
> delay is done very nicely by dummynet in FreeBSD or
> (untested by me ) with NS3 in linux
>
> i don't understand simulating jitter, but, one can always use
> "delay + random number"
>
> > It would be good if that wanted tool is not only able to generate
> > different traffic patterns
>
> if you want to play with the headers ... that'd imply playing with
> nmap/hping3/socat and dozens of other equivalent apps
>
> if you're just trying to flood the wire ... nc/socat/iperf etc
>
> > but also is able to collect different traffic
> > metrics over time. So that it is possible to create catchy plots. :)
>
> "what metrics" you want to collect and how to you want to see it
> would dictate which apps you'd be using
> - tcp queue/buffers
> - dropped packets
> - delays
> - retries
> - udp vs tcp vs icmp vs ...
> - stuff ...
>
> xmit/recv buffers in the hardware, default buffers in the OS and
> buffers in the software apps must all be tuned to the same gigE
> or 10gigE speeds otherwise, whacky stuff will happen
>
> for "catchy plots", you'd want gnuplot so you can (infinitely) zoom in
> into the section you want to see dot-by-dot
>
> for big picture ... netstat, ntop, (not much info) mrtg, etc, etc
>
> big list of apps
> Packet-Craft.net/Apps
>
> > Any hints or links would be greatly appreciated.
>
> if you're a proficient python'er, you'd probably like scapy
> which can do everything you'd need to customize any packet
>
> magic pixie dust
> alvin
> #
> # Packet-Craft.net/Apps
> #
>
>
