Deployment of Distributed Applications in a High Performance Computer Environment

Abstract

Heterogeneous computing (HC) environment is the coordinated use of different types of machines, networks, and interfaces to maximize the ability to solve computationally intensive problems. Usually theses applications consist of various components that have different computational requirements. As machine architectures become more advanced, the extent to which a given task can exploit a given architectural feature depends on how well the task’s computational requirements match the machine’s advanced capabilities. This variation in tasks needs as well as machine capabilities has created a very strong need for developing Mapping/scheduling techniques especially for the HC community. In fact, the applicability and strength of HC systems are derived from their ability to match computing needs to appropriate resources. The mapping problem can be stated shortly as: deciding on which task should be moved to where and when, to optimize some system performance criteria. Mapping problems are known to be NP-Complete except under a few special situations. The existing heuristics for mapping tasks in HC systems works either statically or dynamically. This distinction is based on the time at which the mapping decisions are made. In contrast to static techniques where the complete set of tasks to be mapped is known a priori and the mapping is done off-line, in dynamic mapping methods the machine allocation process is done at run time. Although the principal advantage of the static mapping is its simplicity, it fails to adjust to changes in the system state. A dynamic scheme is needed because the arrival times of the tasks may be random and some machines in the suite may go off-line and new machines may come on-line.