E-DPSIW-FCA: ENERGY AWARE FCA-BASED DATA PLACEMENT STRATEGY FOR INTENSIVE WORKFLOW

Intensive Workflows are composed of large number of complex tasks and require a large amount of data located in different Storage Computing Servers (SC). The data movement between SC causes high communication and data movement cost. In this paper, a data placement strategy based on Formal Concept Analysis approach (E-DPSIW-FCA) is proposed aiming to reduce the data movement, the consumed energy, and the workflow execution cost. FCA allows to group the maximum of data and tasks in an hierarchical structure called lattice concepts. These concepts are mapped to the appropriate SC. The navigation through the hierarchy of concepts is considered as a solution of the case when the data group size exceeds the SC storage capacity. The simulations results show that E-DPSIW-FCA can achieve better results than the K-means [4] and genetic algorithm [14] based approaches.