For middleware compilation log4j library should be added
The middleware can be launched with the start.sh script, with passing there necessary arguments.
The experimental result are situated in zip archive.
Part1 : 1 server
Part1_repeat: 1 server, try less clients
Part2 : 3 servers
Part3: 1 middleware, 1 server
Part4: 1 middleware, 3 servers
Part5: 2 middlewares, 1 server
Part6: 2 middlewares, 3 servers
test_ping - this was where different number of clients were tested in order to determine the ping time from client to middleware.
mw_1.log
mw_2.log
- these are middleware logs from different experiments
stat_*.csv - statistics from dstat
The photos attached are just screenshots when each part for a signle value size ended in order to search more easily for necessary entries in dstat logs.

There are also logs, which fully automates the analysis of the data.
They are in the form of the iPython nothebook, as it was the most convenietn way to work with plots, text and code simultaneously.
They are named again according to which part they refer to.
Part1 - Part6*.ipynb for the plots
Equation*.ipynb - for 2k models
Queueing theory - for Queueing theory computations.

All script contain custom passes, so in odrder to run them the corret passes should be provided. In general these are pathes to experiments and logs received from them

Please, descard scripts with "first middleware" and "second middleware" naming

There is a bunch of scripts to run each experiment.
They are named according to the experiment and are situated in the root folder.
The sript whole experiment launches them all.