Open Access Open Access  Restricted Access Subscription Access

ORR: Optimal Resource Reservation In Grid Computing Environments


Affiliations
1 Department of Computer Science, Government Arts College(W), Salem – 8, TN, India
2 Department of Computer Science, MVM Government Arts College (W), Dindigul, TN, India
 

Objectives: The complex problems are solved by sharing the resources in the grid environment. The proposed ORR approach ensures the availability of resources when it is required in the grid computing environment. The ORR approach also strives to reduce the process switching which is high in TARR (Time-Slice based Resource Reservation). Methodology: The ORR maintains slice queue in which the free time slots are entered. If the free time slices are taken as such then there increase in overhead on process switching. So, the selectslicequeue maintains the slices in best-fit approach based on the current request. In this paper, samples of ten jobs in two scenarios are considered. It is compared with the existing FCFS (First Come First Serve) basis and TARR approach. Findings: In the existing FCFS approach, there is resource reservation denial because of non-availability of resources at the time slot requested. Though the resources are reserved in time-slice basis in TARR based on the defer time, the number of process switching is high. In this case even the small chunk of time-slice is used. The increased process switching increases the overhead on process suspension and resumption. Hence the introduction of bestfit strategy in ORR approach reduces the reservation denial of FCFS and also reduces the overhead on process suspension and process resumption as in TARR. Applications: The ORR scheme can be incorporated in the existing grid environment as Globus, Nimrod G, Legion etc., New agreed virtual organizations can also make use of the ORR scheme.

Keywords

Context Switching, Grid Computing, Optimal Resource Reservation, Resource Management, Slice Queue, TLB.
User

Abstract Views: 136

PDF Views: 0




  • ORR: Optimal Resource Reservation In Grid Computing Environments

Abstract Views: 136  |  PDF Views: 0

Authors

S. Nirmala Devi
Department of Computer Science, Government Arts College(W), Salem – 8, TN, India
A. Pethalakshmi
Department of Computer Science, MVM Government Arts College (W), Dindigul, TN, India

Abstract


Objectives: The complex problems are solved by sharing the resources in the grid environment. The proposed ORR approach ensures the availability of resources when it is required in the grid computing environment. The ORR approach also strives to reduce the process switching which is high in TARR (Time-Slice based Resource Reservation). Methodology: The ORR maintains slice queue in which the free time slots are entered. If the free time slices are taken as such then there increase in overhead on process switching. So, the selectslicequeue maintains the slices in best-fit approach based on the current request. In this paper, samples of ten jobs in two scenarios are considered. It is compared with the existing FCFS (First Come First Serve) basis and TARR approach. Findings: In the existing FCFS approach, there is resource reservation denial because of non-availability of resources at the time slot requested. Though the resources are reserved in time-slice basis in TARR based on the defer time, the number of process switching is high. In this case even the small chunk of time-slice is used. The increased process switching increases the overhead on process suspension and resumption. Hence the introduction of bestfit strategy in ORR approach reduces the reservation denial of FCFS and also reduces the overhead on process suspension and process resumption as in TARR. Applications: The ORR scheme can be incorporated in the existing grid environment as Globus, Nimrod G, Legion etc., New agreed virtual organizations can also make use of the ORR scheme.

Keywords


Context Switching, Grid Computing, Optimal Resource Reservation, Resource Management, Slice Queue, TLB.



DOI: https://doi.org/10.17485/ijst%2F2016%2Fv9i48%2F140224