=========SUMMARY========= - Running of PAES method and Exhaustive method. Both use callcheddar to perform scheduling and security analysis of solutions #Input of Call cheddar - Hyperperiod of the system (we use hyperperiod = max(offset of partitions) + number of partitions * PPCM(tasks periods) - A xml file containing a system (set of processors, tasks, dependencies ...) #Output of Call cheddar - Output .txt: the result of the schedulability and security analysis of the candidate solution as well as values of potential fitness functions - Output2 .txt: This file contains the result of scheduling simulation of the candidate solution #Input of PAES method - Number of tasks in the initial system - A xml file containing a system (set of processors, tasks, dependencies ...): the initial system - Maximum number of partitions - Number of iterations of the PAES method (generation of candidate solutions, their evaluation, archive updating, etc...) - Scheduling policy - PAES strategy : global(compare the mutate solution to the archive) or local (compare mutate solution to the current solution) - List of objectives fitness #Output of PAES method - Execution_trace.txt : shows the list of non-dominated solutions and the values of their objectives fitness - Front.dat : show on each line, the list of the values of objectives fitness of a non-dominated solution - PAES_execution_iter.dat : shows the values of objectives fitness of each feasible candidate solution, the number of rejected solution during archiving process, the number of rejected solutions during mutations(non feasible solutions) - runtime.txt : show the runtime of the PAES method - Solution"i".xmlv3: representes the "i"th non-dominated solutions =========INSTRUCTION FOR PAES METHOD========= - cd {Cheddar Directory}/trunk/src - source ../compilelinux64.bash - make callCheddar2 - make F2T_paes_method2 INSTRUCTION FOR PAES METHOD WITH AN INITIAL SYSTEM MODEL - ./F2T_paes_method2 + -n :Number of the tasks in the initial system + -i :input initial system model + -nb_partitions : Maximum number of partitions + -iter : number of iterations + -sched : Scheduling policy + -select : PAES strategy + -fitness : List of fitness + -DW_capacity : Capacity of downgraders Example: Running PAES on the ROSACE study case; hyperperiod = 4250; time_unit of partition = duration of partition = 250 ./F2T_paes_method2 -n 15 -i test_Rosace_initial -nb_partitions 2 -iter 5000 -sched HOP -select global -fitness "f1 f2 f3" -DW_capacity 166 INSTRUCTION FOR PAES METHOD WITHOUT AN INITIAL SYSTEM MODEL (We generate an initial system model by using Unifast) - ./F2T_paes_method2 + -n :Number of the tasks to generate + -nb_partitions : Maximum number of partitions + -iter : number of iterations + -sched : Scheduling policy + -select : PAES strategy + -fitness : List of fitness + -u : Total CPU utilization + -n_diff_periods : number of maximum different periods + -DW_capacity : Capacity of downgraders Example: Running PAES on the ROSACE study case; hyperperiod = 4250; time_unit of partition = duration of partition = 250 ./F2T_paes_method2 -n 15 -nb_partitions 2 -iter 2 -sched HOP -select global -fitness "f1 f2 f3" -u 90 -n_diff_periods 2 -DW_capacity 166 #Input of Exhaustive method - Number of tasks in the initial system - A xml file containing a system (set of processors, tasks, dependencies ...): the initial system - Maximum number of partitions - Scheduling policy - List of objectives fitness #Output of Exhaustive method - solutions_exact_front.txt : shows the list of non-dominated solutions and the values of their objectives fitness - front_optimal.dat : shows on each line, the list of the values of objectives fitness of a non-dominated solution - Exhaustive_Execution_iter.dat : shows the values of objectives fitness of each generated solution, the number of rejected solution during archiving process - runtime_exhaustive_method.txt : shows the runtime of the Exhaustive method - Solution"i".xmlv3: representes the "i"th non-dominated solutions =========INSTRUCTION EXHAUSTIVE METHOD========= - cd {Cheddar Directory}/trunk/src - source ../compilelinux64.bash - make callCheddar2 - make F2T_exhaustive_method2 INSTRUCTION FOR EXHAUSTIVE METHOD WITH AN INITIAL SYSTEM MODEL - ./F2T_exhaustive_method2 + -n :Number of the tasks in the initial system + -i :input initial system model + -nb_partitions : Maximum number of partitions + -sched : Scheduling policy + -fitness : List of fitness + -DW_capacity : Capacity of downgraders Example: Running Exhaustive on the ROSACE study case; hyperperiod = 4250; time_unit of partition = duration of partition = 250 ./F2T_exhaustive_method2 -n 15 -i test_Rosace_initial -nb_partitions 2 -sched HOP -fitness "f1 f2 f3" -DW_capacity 166 INSTRUCTION FOR EXHAUSTIVE METHOD WITHOUT AN INITIAL SYSTEM MODEL (We generate an initial system model by using Unifast) - ./F2T_exhaustive_method2 + -n :Number of the tasks in the initial system + -nb_partitions : Maximum number of partitions + -sched : Scheduling policy + -fitness : List of fitness + -u : Total CPU utilization + -n_diff_periods : number of maximum different periods + -DW_capacity : Capacity of downgraders Example: Running Exhaustive on the ROSACE study case; hyperperiod = 4250; time_unit of partition = duration of partition = 250 ./F2T_exhaustive_method2 -n 15 -nb_partitions 2 -sched HOP -fitness "f1 f2 f3" -u 90 -n_diff_periods 2 -DW_capacity 166