GT4 Introduction
In the past few years the numbers of PCs have been increasing incessantly and so has the need for ideal power. A growing appetite for computational power in scientific research and computational discovery has resulted in conditions that are favorable for the evolution of computational intermediaries. One of the most significant intermediaries is Grid computing that harnesses the ideal power of the PCs. The term grid computing defined as:
“A type of parallel and distributed system that enables the sharing, selection, and aggregation of geographically distributed autonomous resources dynamically at runtime depending on their availability, capability, performance, cost, and users quality-of- service requirements.”
To implement the Grid environment we need middleware that will help us to share, select and aggregate the distributed resources. There are many middleware’s like SUN N1GE6, Condor, Alchemi, Legion etc. are available for this purpose. Here we are discussing the Globus GT4 middleware. The open source Globus® Toolkit is a fundamental enabling technology for the "Grid," letting people share computing power, databases, and other tools securely online across corporate, institutional, and geographic boundaries without sacrificing local autonomy. The toolkit includes software services and libraries for resource monitoring, discovery, and management, plus security and file management.
From version 1.0 in 1998 to the 2.0 release in 2002 and now the latest 4.0 version based on new open-standard Grid services, the Globus Toolkit has evolved rapidly into what The New York Times called "the de facto standard" for Grid computing. The GT4 architecture consists of following components:
•Security
The GT4 Security components collectively referred to as Grid Security Infrastructure (GSI), facilitate secure communications and the application of uniform policies across distinct systems.
- Delegation: The toolkit includes a service that delegates credentials to a container.
- Community Authorization: Virtual organizations can use the Community Authorization Service (CAS) to manage authorization policies for VO resources.
- Credential Management: This component includes SimpleCA, a simple Certificate Authority for users without access to a full-blown CA, and MyProxy, an online credential repository.
•Data Management
The Data Management components provide for the discovery, transfer and access of large data.
- GridFTP: This component includes a fully functional GridFTP server, and several client-side utilities. The GridFTP protocol is specially optimized to transfer large amounts of data between hosts.
- RFT: The Reliable File Transfer service is a WSRF-enabled service that uses GridFTP internally to move large amounts of data. It provides several interesting features over GridFTP, such as the possibility of resuming interrupted transfers.
- Replica Location: The Replica Location Service (RLS) allows users to keep track of where different replicas of a dataset are located in a virtual organization.
- Data Replication: The Data Replication Service (DRS) uses RLS and RFT to guarantee that local copies of replicas are available to the hosts that need them.
- OGSA-DAI: OGSA Data Access and Integration provides a framework to access and integrate datasets on a Grid which might be available in different formats (plain text files, databases, XML files, etc.).
Execution Management components deal with the deployment, scheduling and monitoring of executable programs, referred to as jobs.
- Grid Resource Allocation & Management (GRAM): GRAM is the heart of GT Execution Management, providing services to deploy and monitor jobs on a Grid.
- Community Scheduler Framework (CSF): This component provides a single interface to different resource schedulers such as PBS, Condor, LSF and SGE.
- Workspace Management: A new component in the toolkit that allows users to dynamically create and manage workspaces on remote hosts.
- Grid Telecontrol Protocol: This component provides a WSRF-enabled service interface for telecontrol (control of remote instruments)
• Information Services
Information Services, commonly referred to as the Monitoring and Discovery System (MDS), includes a set of components to monitor and discover resources in a virtual organization.
- Index Service: This component is used to aggregate resources of interest to a VO.
- Trigger Service: Like the Index service, the Trigger service also collects data from resources, but is configured to perform certain actions based on that data.
- WebMDS: Provides a web browser-based view of data collected by GT4 aggregator services.
The Common Runtime components provide a set of fundamental libraries and tools for hosting existing services as well as developing new services.
- C Runtime: Includes tools, libraries and a WS hosting environment for C developers.
- Python Runtime: Includes tools, client libraries and a WS hosting environment for Python developers.
- Java Runtime: Includes tools, libraries and a service hosting environment for Java developers.
Ian Foster, Carl Kesselman, and Steven Tuecke, The anatomy of the Grid, John Wiley and Sons(2003)
Borja Sotomayor and Lisa Childers, Globus Toolkit 4: Programming Java Services, Elsevier
1 comment:
Nice work i really enjoyed reading this one!!!
Post a Comment