==============================
|   Data Sets structure:     |
==============================


Data sets may be stored in directories called DataSets_1 .. DataSets_n
Each such directory should contain 1 file describing the machines: machines.rnd
Also it should include files describing jobs: jobs0001.rnd .. jobsXXXX.rnd
Each jobsXXXX.rnd file contains information about all jobs for one simulation run.
This allows to run multiple tests using different data sets (i.e. jobs) when using the same environment (i.e. machines + algorithms).
Jobs within one directory (DataSets_x) are supposed to have same average interarrival time - but you can modify it 
to fit your requirements, it is simply useful to keep your test tidy :-)
After you generate the data sets just modify the ExperimentSetup.java class so your simulation will use all data sets:
line 61: for(int folder = 1; folder <=XXX ; folder ++){  // XXX = number of directories
line 63: for(int pass_count = 1; pass_count <=YYY; pass_count ++){ // YYY = number of jobsXXXX.rnd files in one directory

Structure of jobsXXXX.rnd file:
-------------------------------
1st line - number of jobs in the file.
2nd.. nth line = each line represents one job, job parameters are separated by space.
The structure is following:
job_id, number_of_CPUs_requested, RAM_required_(not used now), exec_time_on_specific_machine, arrival_time, specific_machine_speed, deadline_(2147483647=no_deadline), < >, job_input_filesize, job_ouptut_filesize 

Example:
2
1 2 1024 508 2 1076 2147483647 < > 110 50
2 5 64 394 7 2132 2242 < > 732 111

Here we have 2 jobs, first requires 2 CPUs and 1024MB of RAM (not used now), arrive at time 2 (seconds) into system and has no deadline.
Second job requires 5 CPUs and 64MB of RAM (not used now), arrive at time 7 (seconds) into system and has deadline setup on time = 2242 seconds.
1st job execution length can be computed as 508*1076, 2nd job has exec. length = 394*2132 (time on specific machine * specific machine speed)

Structure of machines.rnd file:
-------------------------------
1st .. nth line = each line represents one machine
The structure is following:
CPU_count RAM_size_in_MB Speed_in_MIPS

Example for 5 machines:

5
16 1024 300
8 1024 500
2 512 250
4 512 450
16 1024 320

---------------------------------------------------------------------------------------------
Alea Simulator is distributed under GNU LESSER GENERAL PUBLIC LICENSE
- see lgpl.txt file to read it.
- Original GridSim is protected by GPL licence.