QoS Guided Min-Min Heuristic for Grid Task Scheduling

Task scheduling is an integrated component of computing.With the emergence of Grid and ubiquitous computing, new challengesappear in task scheduling based on properties such as security, qualityof service, and lack of central control within distributedadministrative domains. A Grid task scheduling framework must be ableto deal with these issues. One of the goals of Grid task scheduling isto achieve high system throughput while matching applications withthe available computing resources. This matching of resources in anon-deterministically shared heterogeneous environment leads toconcerns over Quality of Service (QoS). In this paper anovel QoS guided task scheduling algorithm for Grid computing isintroduced. The proposed novel algorithm is based on a general adaptivescheduling heuristics that includes QoS guidance. The algorithm isevaluated within a simulated Grid environment. The experimental resultsshow that the new QoS guided Min-Min heuristic can lead to significantperformance gain for a variety of applications. The approach is comparedwith others based on the quality of the prediction formulated byinaccurate information.