A novel deadline-constrained scheduling to preserve data privacy in hybrid Cloud

Abstract

Cloud computing is always facing privacy and security challenges and this makes the enterprises and the research organizations hesitate to utilize the Cloud technology. Therefore without appropriate solutions to maintain security and privacy of sensitive information in the cloud environment, the potential ability of cloud, which might change the computation pattern, can lead to a vast breakdown in the information technology business. In this paper, a new scheduling model is proposed for workflow applications, which uses the hybrid cloud architecture in order to maintain privacy of the users' sensitive and private tasks and information. The proposed method schedules the sensitive tasks on the private cloud, which is the property and under control of the organization. If a user defined deadline is required to be met, the scheduler is able to schedule other tasks of the workflow on pay-per-use resources of the public cloud. In addition, the scheduler tries to find the schedule map with minimum cost. Experimental results on both scientific and randomly generated workflows show higher performance in the cost and success rate than the comparing methods.