An efficient web cache replacement policy

Several replacement policies for web caches have been proposed and studied extensively in the literature. Different replacement policies perform better in terms of (i) the number of objects found in the cache (cache hit), (ii) the network traffic avoided by fetching the referenced object from the cache, or (iii) the savings in response time. In this paper, we propose a simple and efficient replacement policy (hereafter known as SE) which improves all three performance measures. Trace-driven simulations were done to evaluate the performance of SE. We compare SE with two widely used and efficient replacement policies, namely Least Recently Used (LRU) and Least Unified Value (LUV) algorithms. Our results show that SE performs at least as well as, if not better than, both these replacement policies. Unlike various other replacement policies proposed in literature, our SE policy does not require parameter tuning or a-priori trace analysis and has an efficient and simple implementation that can be incorporated in any existing proxy server or web server with ease.