Grid Computing Infrastructure Architecture Document

4                   Use Case Views
This view presents the users perception of the functionality provided by portal implementations of the Grid and underlying Kernel's technical standards. These use cases were synthesized from [Smothers 2002] but do not include all descriptive text.

General

Applications should be able to sent from any location

Job submission:

Should utilize local batch scheduler

Monitoring

            Global

              Network availability

            Static info      

CPU characteristics, OS, disk

            Dynamic info

Queue info, load

Software

Operating system: Modified Linux kernel with services for computational grid

Computational software, Application software and databases

Agents:

Uses the monitoring information to make decisions

 Algorithm may be application dependent

Data Replication:

            Data and application logic is replicated on different nodes

            A database is maintained for the overall status of the applications

User interfaces

            Job submission interface

                                                Provide a job descriptor file for the job to be submitted

            Interface for users to monitor job status.

 

next      previous