Etl Workflow Optimization And Performance Analysis

In recent years, the data warehouse has been developed to support business decisions, or even a new height to support the business partners and customers. The application of a new generation data warehouse can not only improve the formation of enterprise strategy, but also much more important develop the decision-making and execution ability of the strategic. ETL (extract, transform, and load) is the most important part of the data warehouse system development. The availability of the data warehouse is depended on the correctness of the ETL process. This paper studies the optimizations of ETL processes and analyses the performance of the optimizations to adopt strictly synchronized stochastic Petri nets.First of all, this paper introduces an ETL activity model, analyses the factors which affect the runtime of the ETL activities. The paper sets up a theoretical framework for the problem by modeling it as a state space search problem with which each state graph represents a particular design of the workflow as a graph, equivalents workflows which are produced from state transitions, and the state space is fabricated through a set of correct state transitions, then choose the minimization of the execution cost of the ETL workflow as the best one.Then the paper imports the state space search algorithm, realizes the optimization on ETL workflow by the optimization on the greedy and heuristic search algorithms, and demonstrates the efficiency of the approach through a set of experimental results, which provides the very good reference data for the control of the ETL implementation.Finally, the paper uses strictly synchronized stochastic Petri nets to describe the ETL workflow model, and conducts performance estimation on ETL workflow based on the previous experiments, which improves the validity of the ETL workflow optimization again.