Task Ranking and Allocation Heuristics for Efficient Workflow Schedules

摘要

Task ranking and allocation are two major steps in list-based workflow scheduling. This paper explores various possibilities, evaluates recent approaches in the literature, and proposes several new task ranking and allocation heuristics. A series of simulation experiments have been conducted to evaluate the proposed heuristics. Experimental results indicate that effectiveness of task ranking and allocation heuristics largely depends on the characteristics of workflows to be scheduled, and our new scheduling heuristics can outperform previous methods when dealing with workflows of high CCR (Communication-to-Computation Ratio) values.