Performance Analyse And Investigation For Automatic Self-Allocating Arithmetic And QoS Control Of ASAS Cluster

With the development of the Internet, Web is being more and more widely used, and the requirement for its function as well as performance is getting higher and higher. Meanwhile, a huge potential market is found for the application of Web cluster, the technology of which advances with each passing day. As one of the key technologies in Web cluster, loading balance methods have direct influence on the performance of the cluster, thus it is of great importance to study various scheduling algorithms in qualitative way. Another factor for the successful application of cluster technology is QoS control technology, the proper use of which makes the customers assuredly enjoy timely and reliable service.The main tasks done by the author in this paper are as follows:<1> On the basis of Petri net theory, research is made on the performance of scheduling algorithms in the loading balance of web cluster. This research is further used as the bedrock of the author's analysis on the automatic self-allocating strategy of ASAS. Besides, the loading balance scheduling algorithm based on Threshold-cycle is brought forward, which has already been proved by experiments to bring greater efficiency than other scheduling algorithms.<2> Model for QoS control technology in the cluster is built on the basis of Petri net theory. Comparison among several control algorithms is made by the author through analyzing their respective characteristics and performance, and final conclusion is summarized which bears guiding significance on practical application.