Experimental Analysis on CTT-SP Algorithm for Intermediate Data Storage in Scientific Workflow Systems

Scientific workflows (SWFs) play significant roles in scientific researches and engineering simulations, which are often data intensive and of complex data independency. The storage of the intermediate datasets has important impacts on the performance of SWF systems and, is very difficult to be determined. This paper analyzes the CTT-SP algorithm Yuan2011 that proposed for such problems to provide users the storage strategies aiming at minimizing the total cost. The advantages and disadvantages of the CTT-SP algorithm are discussed. After that, experiments are designed aiming to verify the analysis of the algorithm. At last, the experimental results demonstrate the analysis of the CTT-SP algorithm.