Batch Scheduling based on QOS in HPC - A Survey

High performance distributed applications require high performance and Quality Of Service (QOS). Because of its complex runtime environment and ?with these requirements, it ?makes such applications extremely challenging. Performance evaluation and prediction of these applications in computing environment is a very difficult task. High performance clusters gains the relevance, as a common place for such computational intensive applications. Such applications typically run on a set of heterogeneous machines with dynamically varying loads. These heterogeneous machines are connected by heterogeneous networks possibly supporting a wide variety of communication protocols. The basic problem in parallel computing is how to execute a parallel program on a collection of heterogeneous processors, that is processors with different and possibly changing speeds. Many scheduling algorithms are designed to execute a job efficiently in a parallel computing environment. The goal of this paper is to study on recent batch scheduling methods used for resource allocation across multiple nodes in multiple ways and the impact of these algorithms. This paper presents a review of the literature on batch-processing problems by analysing QOS metrics in high performance computing environment. Herein we discuss about the area and the significance of scheduling, based on papers reviewed prevoiusly.This paper proposes an approach for batch scheduling of jobs by considering certain QOS metrics.