EZ-Grid is an ongoing project at the Department of Computer Science, University of Houston. The aim is to design and implement a resource brokerage system coupled with user interfaces and robust information objects for multi-site grid computing. We use Globus tools for grid services and develop the above software tools for making resource selections and job submission to achieve the time and/or cost constraints specified by the user.

Given the current status in Grid Computing, selecting and using the grid resources requires the user to have explicit knowledge about the middleware, grid resources required and how to express the resource requirements. This task could be complicated when the user desires the best performance for his job execution in terms of either time and/or cost. Thus, using the grid through appropriate services together with knowledge of the resources to obtain the best or at least acceptable performance for job executions is an extremely complicated task. The solution is to make the usage of the Grid easier and transparent for the user. This can be achieved by developing easy-to-use interfaces coupled with brokerage systems to assist the resource selection and job execution process. Such tools would relieve the end user of making resource selections, specifications and job executions by themselves.

EZGrid system offers a high-level job submission interface separating the internals of middleware from the user, a broker kernel equipped with sophisticated scheduling algorithm to offer best machine choices and a job monitor for monitoring the status and failures, if any, of submitted jobs

 Architecture