Scheduling Data-Driven Workflows in Multi-CloudEnvironment

Nowadays, cloud computing and otherdistributed computing systems have been developed tosupport various types of workflows in applications. Due tothe restrictions onthe use ofone cloud provider, the conceptof multiple clouds as been proposed.Inmultipleclouds,schedulingworkflowswithlarge amounts ofdata is a wellknownNP-Hard problem. The existing schedulingalgorithms have not paid attention to the data dependencyissues and their importance in scheduling criteria such astime and cost. In this paper, we propose a communicationbasedalgorithm for workflows with huge volumes of datain a multi-cloud environment. The proposed algorithmchanges the definition of the Partial Critical Paths(PCP) tominimize the cost of workflow executionwhile meeting auser defined deadline.

کلیدواژه ها:

Cloud Computing ، Multi-cloud ، Workflow Scheduling ، Data Dependency ، Communication