[tor-dev] Simulating a slow connection

[Rob Jansen]

> Date: Thu, 26 Jan 2012 16:39:34 +0000
> From: Steven Murdoch <Steven.Murdoch at cl.cam.ac.uk>
> To: tor-dev at lists.torproject.org
> Subject: Re: [tor-dev] Simulating a slow connection
> Message-ID: <CAE309A8-034E-4E6A-B05C-6B9B3EA1CDBE at cl.cam.ac.uk>
> Content-Type: text/plain; charset=us-ascii
> 
> Hi Adam,
> 
> On 20 Jan 2012, at 10:55, Adam Katz wrote:
> 
>> Well, I myself didn't have anything specific in mind but i have some
>> experience with the linux tc utility as well as with generating
>> realistic background traffic. I was wondering whether I could help on
>> any of the existing projects or help establish a new one.
> 
> I think Nick's comments summarized the current state of thinking. ExperimenTor and Shadow are the best Tor simulators to use for this project. The big missing pieces are:
> - an automated framework for setting up experiments with slow Internet connections with ExperimenTor and Shadow, then collecting and summarizing results
> - Tools for generating realistic link characteristics (delay and packet dropping), and for collecting data on the link properties found in particular locations
> 
> Steven.
> 

As Steven stated, this would be very easy to explore with Shadow. The
network topology is passed in as an XML file: node properties include
bandwidth up/down and link properties include latency, jitter, and
packetloss.

I already have some python scripts to generate topologies, and would be
happy to share them once you have realistic measurements/values for the
slow links you'd like to explore. I'd also be happy to explain more
about how Shadow works or the structure of the XML file.

Rob

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 262 bytes
Desc: OpenPGP digital signature
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20120202/16528f5c/attachment.pgp>