gridsim:java-based modelling and simulation of deadline and budget-based scheduling for grid...
TRANSCRIPT
GridSim:“Java-based Modelling and
Simulation of Deadline and Budget-based Scheduling for Grid
Computing”
Rajkumar Buyya and Manzur Murshed
Monash University, Melbourne, Australia
www.buyya.com/ecogrid
www.gridcomputing.com
Simulation Parameters
Resources Cost : cheapest: 10, Expensive: 20 (normal distribution for costing. Speed: 0.5 to 1.5 (1, standard machine. normal distribution for
speed). Users:
Users job contains 20 tasks with variation of +/-2 with random submission.
Users submitted jobs only after completion of previous job. Jobs = 20 tasks
Each task takes 50units. Heterogeneous tasks (future)
Simulation Time = 7*60*60 units (approx.7hours). As the number of users grows, the probability of getting at least
one resource per user, throughout the deadline, decreases. This low probability demands high (>> 1) D_Factor and B_Factor
in order to achieve very high job completion rate.
D-Factor
MINMAX
MIN
MIN
MAX
Job_TimeJob_Time
Job_TimeDeadlineFactorD
Job_Time
Job_Time
_
priority highest theresourcefastest thegiving
parallel,in tasks, theall process toTime
resourceslowest theusing
serially, tasks, theall process toTime
Any job with D_Factor < 0 would never be completed
As long as some resources are available throughout the deadline, any job with D_Factor 1 would always be completed
B-Factor
MINMAX
MIN
MIN
MAX
Job_CostJob_Cost
Job_CostBudgetFactorB
deadline
Job_Cost
deadline
Job_Cost
_
priorityhighest theresource
cheapest thegiving , within
parallelin tasks, theall process Cost to
priorityhighest theresource
costliest thegiving , within
parallelin tasks, theall process Cost to
• Any job with B_Factor < 0 would never be completed• As long as some resources are available throughout the
deadline, any job with B_Factor 1 would always be completed
Job Completion & Time Optimise
50 90
130
170
210 50
17075
80
85
90
95
100
Job completion
rate (%)
B-factor (%)
D-factor (%)
Users=21, Resources=25, Optimization: TIME
Time Utilisation & Time Optimise
50 90
130
170
210 50
15001020304050607080
Time Utilization (%)
D-factor (%)
B-factor (%)
Users=21, Resources=25, Optimization: TIME
Budget Utilisation & Time Optimise
50 90
130
170
210 50
1700
10
20
30
40
50
60
Budget Utilization (%)
B-factor (%)
D-factor (%)
Users=21, Resources=25, Optimization: TIME
Job Completion & Cost Optimise
50 90
130
170
210 50
1700
20
40
60
80
100
Job completion
rate (%)
B-factor (%)
D-factor (%)
Users=21, Resources=25, Optimization: COST
Time Utilisation & Cost Optimise
50 90
130
170
210 50
1500
20
40
60
80
100
Time Utilization (%)
B-factor (%)
D-factor (%)
Users=21, Resources=25, Optimization: COST
Budget Utilisation & Cost Optimise
50 90
130
170
210
50
1700
5
10
15
20
Budget Utilization (%)
B-factor (%)
D-factor (%)
Users=21, Resources=25, Optimization: COST
Job Completion for Optimise Time
1 5 9
13 175
150
20
40
60
80
100
Job completion
rate (%)
Users
Resources
B-factor = D-factor = 110%, Optimization: TIME
Time Utilisation for Optimise Time
5 9
13 17 21
1
110
10
20
30
40
50
60
70
Time Utilization (%)
Resources
Users
B-factor = D-factor = 110%, Optimization: TIME
Budget Utilisation for Optimise Time
1 5 9
13 175
150
5
10
15
20
25
30
Budget Utilization (%)
Users
Resources
B-factor = D-factor = 110%, Optimization: TIME
Job Completion for Optimise Cost
1 5 9
13 175
150
20
40
60
80
100
Job completion
rate (%)
Users
Resources
B-factor = D-factor = 110%, Optimization: COST
Time Utilisation for Optimise Cost
1 5 9
13 175
150
20
40
60
80
100
Time Utilization (%)
Users
Resources
B-factor = D-factor = 110%, Optimization: COST
Budget Utilisation for Optimise Cost
1 5 9
13 175
150
5
10
15
20
25
Budget Utilization (%)
Users
Resources
B-factor = D-factor = 110%, Optimization: COST