A Randomization Approach for Stochastic Workflow Scheduling in Clouds

Table 1

Stochastic model of the DAG example shown in Figure 1.

(a) Stochastic model of execution times

	Resource 1		Resource 2		Resource 3
	Exp. val.	Std. dev.	Exp. val.	Std. dev.	Exp. val.	Std. dev.
Task 1	24.00	4.00	2.00	0.33	7.00	1.17
Task 2	36.00	6.00	32.00	5.33	49.00	8.17
Task 3	42.00	7.00	49.00	8.17	12.00	2.00
Task 4	12.00	2.00	8.00	1.33	56.00	9.33
Task 5	81.00	13.50	20.00	3.33	16.00	2.67
Task 6	9.00	1.50	30.00	5.00	24.00	4.00
Task 7	48.00	8.00	4.00	0.67	8.00	1.33
Task 8	64.00	10.67	18.00	3.00	42.00	7.00
Task 9	36.00	6.00	16.00	2.67	1.00	0.17
Task 10	54.00	9.00	28.00	4.67	63.00	10.50

(b) Stochastic model of communication times

	Resource 1		Resource 2		Resource 3
	Exp. val.	Std. dev.	Exp. val.	Std. dev.	Exp. val.	Std. dev.
Resource 1	0.00	0	2.00	0.33	8.00	1.33
Resource 2	2.00	0.33	0.00	0.00	4.00	0.67
Resource 3	8.00	1.33	4.00	0.67	0.00	0.00