Network Simulator Page
NS is a network simulation package. Homepage is here.
Installation & Calibration
Installation can be performed on your own home directory. I simply downloaded and installed the 'allinone' package which comprises of the main components of ns. The package was dated June 25, 01 so I would assume that it's quite old. It does however included all sources, so i should be able to update individual components later.
After untar'ing the archive, i run the ./install script. I believe this did all the compiling and building for me... i don't really know!!!
That seemed to go fine, and at the end it told me to include some environmental paths:
I was then told to enter into the ns-2.1b8a directory to run validate. This seems to run a set of tests actually using network protocols on the computer. As such it took quite a while.
After about 20 minutes on a dual Xeon 1.7Ghz with 512Mb, the tests completed. The output was a bit strange saying that all test outputs agree with reference output, but the overall validate report said that 'some tests failed: ./test-all-red.
So i cd'ed into ns-all-in-one-2.1b8a/ns-2.1b8/tcl/test, and rerun test-all-red. This then proceeded to run a series of tests in turn (red, ecn, red_twoway, etc.). There seemed to be a problem with xgraph as it states that it could not execute.
On the execution of the test 'flows_unforced' it states that the output differed from the reference output. It states that i should either compare the output to the reference with the diff command or look at the ns-problems.html page. It appears that the problem arises on Intel Linux machines due to the the implementation of IEE arithmetic. It doesn't sound hopeful that they'll find a fix...
So i carried on and output was also diferrent with flows_combined.
To fix the xgraph problem i included the directory /home/ytl/ns-all-in-one-2.1b8a/bin in the PATH environment. I then proceeded to include all this paths into my .bash_profile so i don't have to bother with it anymore. :)
Rerunning the tests with the paths gave me a nice looking graph:
I dunno what it means, but it looks useful! and there was also another graph underneath this one:
Difficulties in Simulating the Internet. here.
|© 2001-2003, Yee-Ting Li, email: email@example.com,
Tel: +44 (0) 20 7679 1376, Fax: +44 (0) 20 7679 7145
Room D14, High Energy Particle Physics, Dept. of Physics & Astronomy, UCL, Gower St, London, WC1E 6BT