A novel hierarchical cache architecture for WWW servers

This paper presents a novel hierarchical cache architecture to achieve efficient Web access. The cache servers being used today have encountered the problems of lack of efficient collaboration among different levels in the hierarchy. In other words, the highest level of the hierarchical cache servers may easily become the performance bottleneck, which in turn slows down the entire cache system. In the proposed new cache design, we apply an inclusive and exclusive relationship to modify the ICP (Internet cache protocol) and use the recursive concept to build a hierarchical cache architecture to avoid the unnecessary information query existing in current ICP. This implies that the lowest level can fetch the data from the original Web server directly and perform the update procedures recursively to its upper levels. The up-side-down traffic flow, referred to as reverse traffic flow in this paper, can substantially release the server load of the upper levels in the hierarchy. For the purpose of evaluation and comparison, we derive mathematical equations by analyzing the operation procedures step by step. The analytical results have shown that the server load of the highest level in the hierarchy can be reduced almost 50% under the worst-case assumption. Although the lower levels may slightly increase their workload, it does significantly decrease the average object searching time by balancing the loads among different levels.