[131882] in North American Network Operators' Group
Re: Emulating a cellular interface
daemon@ATHENA.MIT.EDU (Adrian Chadd)
Sat Nov 6 21:06:56 2010
Date: Sun, 7 Nov 2010 09:06:48 +0800
From: Adrian Chadd <adrian@creative.net.au>
To: Andy Davidson <andy@nosignal.org>
In-Reply-To: <75424723-658C-4E2C-88E6-D8D12C686080@nosignal.org>
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
On Sat, Nov 06, 2010, Andy Davidson wrote:
> Not withstanding Mikael's comments that it shouldn't be lossy, at times when you want to simulate lossy (and jittery, and shaped, and ....) conditions, the best way I have found to do this is FreeBSD's dummynet :
>
> http://www.freebsd.org/cgi/man.cgi?query=ipfw&sektion=8#TRAFFIC_SHAPER_(DUMMYNET)_CONFIGURATION
And cellular networks are bursty, depending upon (from what I can gather)
how busy the cell is and how many people are currently doing data.
Someone more cellular-oriented should drop in their 2c.
So to be completely accurate, you may way to script some per-node shaping
rules that watch traffic flow and adjust the rules to emulate this.
I recall seeing a few apps that behaved poorly when their UDP data
exchange timed out because my 3G connection was in a "slow" mode and
didn't recover well. It required a background ICMP to keep the damned
session nailed up to "fast". :-)
Adrian
