Replica Selection in the Globus DataGrid

Title: Replica Selection in the Globus DataGrid

Research Question: How can the Globus DataGrid architecture provide a scalable infrastructure for managing distributed data resources with varying quality of service requirements?

Methodology: The researchers designed and implemented a high-level replica selection service that uses information regarding replica location and user preferences to guide selection from among storage replica alternatives. They utilized dynamic information collected using Globus information service capabilities concerning storage system properties to optimize the selection process. They demonstrated the use of Condor's ClassAds resource description and matchmaking mechanism as an efficient tool for representing and matching storage resource capabilities and policies against application requirements.

Results: The researchers presented a basic replica selection service design and showed how dynamic information could be used to improve and optimize the selection process. They demonstrated the use of Condor's ClassAds and matchmaking mechanism as an elegant matching and ranking tool in a storage context.

Implications: The implementation of the replica selection service in the Globus DataGrid architecture provides a scalable solution for managing distributed data resources with varying quality of service requirements. The use of Condor's ClassAds and matchmaking mechanism offers an efficient and flexible method for matching and ranking storage resources against application requirements. This research contributes to the ongoing development of grid computing technologies and their application in scientific research and other data-intensive fields.

Link to Article: https://arxiv.org/abs/0104002v1 Authors: arXiv ID: 0104002v1