Academia.eduAcademia.edu

Figure 3 – uploaded by FOREX Publication

See full PDFdownloadDownload figure
Figure 3: Behavior of Full Checkpoint/Restart Model The fundamental concept is to determine the anticipated amount of time lost as a result of errors, and then make use of this information to recalculate the amount of processing power actually available for resources. Later this new capacity/ processing power is used for making the scheduling decision. The wasted time is calculated as follows:

Figure 3 Behavior of Full Checkpoint/Restart Model The fundamental concept is to determine the anticipated amount of time lost as a result of errors, and then make use of this information to recalculate the amount of processing power actually available for resources. Later this new capacity/ processing power is used for making the scheduling decision. The wasted time is calculated as follows:

Related Figures (7)

The above comparison describes (see Table 1) various scheduling approaches. The strategy proposed in this research is based on failure factors and performance parameters both, which is different from the above methods. Hence the method devised is novel in nature.
The above comparison describes (see Table 1) various scheduling approaches. The strategy proposed in this research is based on failure factors and performance parameters both, which is different from the above methods. Hence the method devised is novel in nature.
Figure 1: Grid Computing Environment [36]
Figure 1: Grid Computing Environment [36]
Figure 2: Mechanism of Checkpoint/Restart as stochastic renewal reward process we reload the saved snapshot. The recovery state has two costs: the time spent loading the most recent snapshot (Rr) and the time spent in re-computing (Tre). As a result of the checkpoint system, time can be saved by not re-computing task from scratch after a failure. Figure 2 and Figure 3 show the behavior of the whole checkpoint/restart paradigm [6].
Figure 2: Mechanism of Checkpoint/Restart as stochastic renewal reward process we reload the saved snapshot. The recovery state has two costs: the time spent loading the most recent snapshot (Rr) and the time spent in re-computing (Tre). As a result of the checkpoint system, time can be saved by not re-computing task from scratch after a failure. Figure 2 and Figure 3 show the behavior of the whole checkpoint/restart paradigm [6].
Figure 4: Flowchart for Proposed Scheduling Algorithm Research Article | Volume 10, Issue 3 | Pages 651-658 | e-ISSN: 2347-470X Various parameters used in algorithm are given as under: PEECMTPICGOUIMIIGH JUULLTICN Ui Flectrical and Electronics Research (IJEER)
Figure 4: Flowchart for Proposed Scheduling Algorithm Research Article | Volume 10, Issue 3 | Pages 651-658 | e-ISSN: 2347-470X Various parameters used in algorithm are given as under: PEECMTPICGOUIMIIGH JUULLTICN Ui Flectrical and Electronics Research (IJEER)
Table 4: Number of Failure and Failure Ratio = Table 5: Average Response Time
Table 4: Number of Failure and Failure Ratio = Table 5: Average Response Time
Research Article | Volume 10, Issue 3 | Pages 651-658 | e-ISSN: 2347-470X PETC TE EIGUIMViiaGs JIVULTIGI Vi Flectrical and Electronics Research (IJEER)
Research Article | Volume 10, Issue 3 | Pages 651-658 | e-ISSN: 2347-470X PETC TE EIGUIMViiaGs JIVULTIGI Vi Flectrical and Electronics Research (IJEER)
Figure 5 demonstrates that the FABS makespan is less than those of SOSA, which means that average time taken by the proposed algorithm is less than average time taken by the existing scheduling algorithm, and hence the efficiency of system is increased. Figure 6 shows that the new algorithm's performance ratio (PR) is more than 1, means that FABS outperforms SOSA method. It signifies that the proposed
Figure 5 demonstrates that the FABS makespan is less than those of SOSA, which means that average time taken by the proposed algorithm is less than average time taken by the existing scheduling algorithm, and hence the efficiency of system is increased. Figure 6 shows that the new algorithm's performance ratio (PR) is more than 1, means that FABS outperforms SOSA method. It signifies that the proposed
Related topics:
Grid Computing
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved