Replica maintenance strategy for data grid
Data Grid is an infrastructure that manages huge amount of data files, and provides intensive computational resources across geographically distributed collaboration.Increasing the performance of such system can be achieved by improving the overall resource usage, which includes network and storage...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universiti Teknikal Malaysia Melaka
2017
|
| Subjects: | |
| Online Access: | http://repo.uum.edu.my/21722/1/JTECE%209%201-2%202017%2047%2051.pdf http://repo.uum.edu.my/21722/ http://journal.utem.edu.my/index.php/jtec/article/view/1650 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Data Grid is an infrastructure that manages huge amount of data files, and provides intensive computational resources across geographically distributed collaboration.Increasing the performance of such system can be achieved by improving the overall resource usage, which includes network and storage resources.Improving network resource usage is achieved by good utilization of network bandwidth that is considered as an important factor affecting job execution time.Meanwhile, improving storage resource usage is achieved by good utilization of storage space usage. Data replication is one of the methods used to improve the performance of data access in distributed systems by replicating multiple copies of data files in the distributed sites.Having distributed the replicas to various locations, they need to be monitored.As a result of dynamic changes in the data grid environment, some of the replicas need to be relocated.In this paper we proposed a maintenance replica placement strategy termed as Unwanted Replica Deletion Strategy (URDS) as a part of Replica maintenance service.The main purpose of the proposed strategy is to find the placement of unwanted replicas to be deleted.OptorSim is used to evaluate the performance of the proposed strategy. The simulation results show that URDS requires less execution time and consumes less network usage and has a best utilization of storage space usage compared to existing approaches. |
|---|