BitTorrent Automatic Command and Analysis Infrastructure

The BitTorrent Automatic Command and Analysis Infrastructure is to be
used for automatic command of BitTorrent clients. BitTorrent clients
will be automatically started, stopped, paused using current framework.

Clients are instrumented in order to output status and verbose logging
messages for subsequent analysis.

Status messages and verbose messages are to be collected, parsed, stored
and analysed. Our goal is to determine specific weaknesses and possible
enhancements in BitTorrent implementation.

We are currently focusing on libtorrent-raster[1] and Tribler[2].

This work is part of the EU FP7 P2P-Next project[3].

[1] http://www.rasterbar.com/products/libtorrent/
[2] http://www.tribler.org/trac
[3] http://www.p2p-next.org/


Contributions from:
 * Adriana Drăghici, adriana008@gmail.com
 * Marius Sandu-Popa, sandupopamarius@gmail.com
 * Răzvan Deaconescu, razvan.deaconescu@cs.pub.ro


smf/
	- Swarm Management Framework
	- scripts and Python programs for automatic deploy and run
	  of BitTorrent clients
	- see smf/README

ppf/
	- Post Processing Framework
	- scripts, Python programs and SQL files for parsing, storing
	  and analysis of BitTorrent client status and verbose log
	  messages
	- see ppf/README

instrumentation/
	- modified BitTorrent clients to be used in the current
	  infrastructure
	- updates/modifications are usually geared towards enabling
	  verbose message output (BitTorrent protocol information) from
	  original client
	- see instrumentation/*/README

old/
	- old scripts and programs from related work
	- see old/README

scripts/
	- various scripts for automation

experiments/
	- experiments/scenarios details, planning and results