A distributed scheduling service using genetic approach

Abstract

Cloud computing is a paradigm to perform distributed processing in newlinean open computing environment. It eliminates the need for high-cost newlinededicated expensive hardware, space and dedicated software to minimal newlineinteraction and management. Recently cloud computing is delivering ondemand newlineservices like software, memory, data, network-bandwidth and other newlineIT related services over the internet. The reliable performance of cloudservices newlineis related primarily to task-scheduling. Scheduling can be done on newlinedifferent levels like job level, infrastructure level or process level. The newlineobjective of this research is to improve task scheduling using a genetic newlineapproach. When the end-user sends the request to the main data-centre for newlinejobs to be computing, the tasks are named. A task means a piece of work or newlineprocess that can be executed within the deadline. Tasks levels are divided into newlinecritical and non-critical, where critical tasks are executed with in the given newlinetime period, and non-critical is considered as low priority work in the cloud newlinesystem. Scheduling dispatches all user tasks provided by the users of the newlinecloud system to the cloud service provider to search for available resources in newlinethe cloud. newlineTask-Scheduling newline