Efficient information retrieval via proxy caching mechanisms within a lossy environment

With the growing demand for high speed availability to online resources, service providers are looking at improving the traditional client server model for information retrieval suffering from scalability and efficiency problems. In this paper we investigate the use of proxy-servers together with caching mechanisms to reduce network load and serve a large number of clients quickly within a lossy environment. A storage system is designed and implemented, coupled with intelligent mechanisms to store either the latest or most popular resources. The performance analysis results show that cached data can be accessed in a much shorter time as the overall system´s data´s transfer time and origin server´s processing load decreases. We have also investigated the correlation of deletion algorithms with different client request patterns and conclude that when each file is not equally popular multiple proxy-servers can reduce the load on the server further.