Author: Xiao Qin 

Address:  Department of Computer Science and Software Engineering, 
Auburn University 
Email: <xqin@auburn.edu>
Date: 2002/7/25 

Version 1.0 
balance-para-tuff

This simulation program was modified from the previous version
written by Yanxia Qu, who was  motivated by reading Allen B. Downey's
simulation program, where several Allen B. Downey's functions are used.

Please cite the folowing publications:
[1] X. Qin, "Design and Analysis of a Load Balancing Strategy in Data Grids," Future Generation Computer Systems: The Int'l Journal of Grid Computing, vol. 23, no. 1, pp. 132-137, Jan. 2007.
[2] X. Qin, "Performance Comparisons of Load Balancing Algorithms for I/O-Intensive Workloads on Clusters,"  Journal of Network and Computer Applications, vol. 31, no. 1, pp. 32-46, January 2008.



Command line:

system trace_file_name LS_policy paging_rate memory_threshold  re_cost fx_cost network disk_paging_cost cpu_speed0 ... cpu_speed5 mem_size0 ... mem_size5

* trace_file_name: e.g. t0_m1 (trace0 with mean memory demand 1MB)
* LS_policy: 0:NO_LS; 1:MEM_RE; 2:CPU_RE; 3:CPU_MEM_HP_RE; 4:CPU_MEM_HT_RE;
                      5:MEM_PM; 6:CPU_PM; 7:CPU_MEM_HP_PM; 8:CPU_MEM_HT_PM;
                     15:IO_RE; 17: IOPAGE_RE; 18: ALL_RE; 19: IOBEST_RE 20:IO_PM 21:IOBEST_PM 22:WAL_RE 23:WAL_PM
		     24:COM_RE
* paging_rate: e.g. 0.1
* memory threshold is 0.4 (40%) of its request memory size for all existing experiments
* re_cost: a remote execution cost
* fx_cost: a fixed remote execution cost in a preemptive migration
* network: Ethernet speed
* disk_paging_cost: e.g. 10
* six machines' cpu speed in MIPS, for example, 100 MIPS or 500 MIPS
* six machines' mem size in MB bytes excluding pre-occupied memory, 
  for example, 64 MB - 16 MB, should input 48.



output files:
  output*   ( slowdown of policy *) : the discription of each data is print out on the screen
  CPU_distr
  mem_distr ( for frequent paging percent )

Trace file format:
1.Arrival_time, 2.arrival_node, 3.duration_time, 4.request_mem_size 5.I/O_rate, 6.Initial I/O node, 7.I/O type, 8. init_data_size, 9. comm_rate, 10. job_name, 11. num of tasks in a job