PROJECT TITLE :

Energy-aware cloud workflow applications scheduling with geo-distributed data

ABSTRACT:

The cost of electricity shifts during the course of the day and varies from one geographic location to another. Workflow applications in the cloud frequently require geo-distributed data, which must then be transmitted between heterogeneous servers located within and between data centers. When trying to optimize the energy cost for scheduling tasks in workflow applications to heterogeneous servers in cloud data centers, one of the greatest challenges comes from the wide range of prices for electricity and the length of time it takes to transmit data. In this piece, we take on the challenge of reducing the total cost of electricity usage within the context of a time-sensitive, energy-conscious workflow scheduling issue in which the data is geographically dispersed across multiple data centers. An algorithm for scheduling is presented here. Applications for workflow are sequenced, deadlines are divided, and tasks are sorted according to various strategies. An adaptive local search method that dynamically balances intensification through the use of neighborhood structures of increasing size is presented as a means of improving solutions during the process of searching for them. Statistics are used to calibrate the values of the components and parameters over a comprehensive data set of random instances. A comparison is made between the proposed algorithm and modified versions of classical algorithms designed to solve problems similar to the one at hand. The effectiveness of the proposal for solving the problem that was considered is demonstrated by the experimental results.