There has been an explosion in the volume of data that is being accessed from\nthe Internet. As a result, the risk of a Web server being inundated with requests\nis ever-present. One approach to reducing the performance degradation\nthat potentially comes from Web server overloading is to employ Web\ncaching where data content is replicated in multiple locations. In this paper,\nwe investigate the use of evolutionary algorithms to dynamically alter partition\nsize in Web caches. We use established modeling techniques to compare\nthe performance of our evolutionary algorithm to that found in statically-\npartitioned systems. Our results indicate that utilizing an evolutionary algorithm\nto dynamically alter partition sizes can lead to performance improvements\nespecially in environments where the relative size of large to\nsmall pages is high.
Loading....