[tcpm] A new performance measurement tool - flowgrind

[Alexander Zimmermann <[e-mail]>]

Hi all,

during our last research projects we ran into some problems with the  
current performance measurement tools like iperf, netperf, and the  
like. Mostly these tools are not able to schedule a bunch of flows  
among a number of nodes in a easy way or report more than throughput  
(e.g., kernel variables like CWND, SSTHRESH).

Amongst others flowgrind implements the following features:

* Distributed architecture
* Sophisticated flow scheduling
* Linux kernel TCP statistics
* Split control and data connection
* Application layer 1-way (IAT) and 2-way (RTT) delay
* Customizable socket options
* Anderson-Darling statistical test
* Monitoring, measuring, and accounting
* Rate-limited flows (Uniform and poisson packet distribution)

For more information, please check our site: http://www.umic-mesh.net/research/flowgrind/

If you interest in flowgrind you can download there tarball or  
directly access the code via
our subversion repository: svn://svn.umic-mesh.net/flowgrind/trunk

Comments and patches are more than welcome!

NB: At moment flowgrind runs only under Linux :-(

Alex

//
// Dipl.-Inform. Alexander Zimmermann
// Department of Computer Science, Informatik 4
// RWTH Aachen University
// Ahornstr. 55, 52056 Aachen, Germany
// phone: [phone], fax: [phone]
// email: [e-mail]
// web: http://www.umic-mesh.net
//