We have proposed and evaluated a static caching policy for the server caching. It has lower management overhead and higher hit rate. Under some statistical assumptions we have shown that static caching gives the highest hit rate. We have also provided empirical comparison results obtained by trace-driven simulations. The simulation results also illustrated that the static caching is more efficient than those analyzed in the literature.
One potential disadvantage of static caching is that, because it infrequently modifies the contents of the cache, it exhibits a long delay in reaction to important events where some document becomes all of a sudden very popular. In this case, the cache has to wait for the next update to bring in this document. Such a situation might occur at the creation of new pages. Thus, a possible remedy could be to put potentially highly required new pages in the cache as soon as possible. This results in a mixed caching policy where part of the cache if static while the other part is dynamic. Such an approach will be investigated in the future studies.