Skip to content

Latest commit

 

History

History
82 lines (44 loc) · 2.2 KB

README.md

File metadata and controls

82 lines (44 loc) · 2.2 KB

Build Status Code Climate Issue Count

FatTreeGenerator

This repository contains an k-ary FAT tree topology generator.

A tcl representing the topology or the given k is genrated on running the python FatTreeGenerator.py script

Example topology:

               (CORE)
              .     .
            .        .
          .           .
        (Agg)         (Agg)
      .     .           ...
     .       .
    (TOR)    (TOR)
   . .         .
  .   .       . .
 .     .     .   .
(0)    (1)  (2)  (3) ...

Numbering for nodes starts at the end host and continues hierarchically i.e. End hosts then TOR then AGG then core switches.

Links are numbered similiarly stating form the node 0.

Summary of the files

cleanup.py

Cleans up the result files in the out/ folder after an experiment.

createUtilGraph.py

Creates the graphs from the result files in the out/ folder

FatTreeGenerator.py

Generates the topology and moves the resulting files in the out/ folder

make_graphs.py

A wrapper script which calls createUtilGraph.py

ratioCalc.py

Calculates the ratio of data plane traffic (TCP) to control plane traffic (eSDN DHT, SDN controller polling traffic)

tcl/

Holds the TCL templates and reference files.

Running the Experiment

Edit the start of the FatTreeGenerator.py file to adjust the topology's K-value and the link bandwidths then run the file using the following command:

python FatTreeGenerator.py

The result will be stored in the out/ folder.

Then run ns2 on the generated TCL file after ajusting the parameters in out.tcl.

cd ../out/

vi out.tcl

ns out.tcl

The simulation will produce its results in the out/ directory.