Master Thesis - ICS

Computer Science Department
Antonis Misargopoulos
A scheduling model consists of a scheduling policy; i.e., a set of rules for
producing schedules, a program model which abstracts the set of programs to be
scheduled, and a performance model which abstracts the behavior of the program on the
underlying system for the purpose of the evaluating the performance potential of
candidate schedules. In addition, the scheduling model utilizes a performance measure
which describes the performance activity to be optimized by the performance model. As a
definition, we can say that high-performance schedulers are software systems which use
scheduling models to predict performance, determine application schedules based on
these models, and take action to implement the resulting schedule. The high-performance
scheduler determines an application schedule; i.e., an assignment of tasks, data and
communication to resources, ordered in time and based on the rules of the scheduling
policy. The goal is to optimize the performance experienced by the application on the
computational Grid.
It is very important to distinguish distributed parallel applications and ‘portable’
single-site applications. Even if an application cannot profit from distribution, the
scheduler may have to locate a computational site and/or co-locate data in a way that
maximizes application performance. For parallel Grid applications, the high-performance
scheduler will need to determine whether performance is optimized by accessing all tasks
to a single site or by distributing the application to multiple sites.
3.2 Developing a Grid Scheduling Model
The development of efficient models for scheduling Grid environments presents a
substantial challenge to researchers and application developers. One approach to
developing Grid models is to compose models from constituent components which reflect
application performance activities. This approach is being taken by a number of
researchers [81, 82, 83, 84]. To illustrate this, consider a simple model which predicts
execution time for a Grid application which executes one task A to completion, and
passes all data to another task B which then executes to completion. Thus, a performance
model for this application is
24