Offline Cache缓存调度算法实验报告.docx

Offline Cache实验报告学院：软件学院 学号：201421031059 姓名：吕吕Question DescriptionConsider the management problem between adjacent levels:1)Main memory with n data items from a set U2)Cache can hold kn items Simplest version with no direct-mapping or other restrictions about where items can be1)Suppose cache is full initially 2)Holds k data items to start with Given a memory request d from U1)If d is stored in the cache we can access it quickly2) If not then we call it a cache miss and (since the cache is full) we must bring it into cache and evict some other data item from the cache which one to evict? Question: Given a sequence D=d1,d2,…,dm of elements from U corresponding to memory requests, Find a sequence of evictions (an eviction schedule) that has as few cache misses as possible。二、Common?cache replacement?algorithm comparison2.1 Traditional cache replacement algorithm2.1.1 FIFO(First Input First Output)Algorithm Idea: The algorithm?eliminate?the page first to enter the?cache?.Advantage: The realization is simple.Disadvantage: Rarely used in practice,?because ?page fault rate of the algorithm?is high.2.1.2 LRU(Least Recently Used)Algorithm Idea: If the data has recently been visited, then in the future probability of access is also higher. Therefore, to eliminate the data most long time without access. LRU algorithm is based on recency .Advantage: The realization of LRU is simple and?the extra expenses are very few. Moreover, LRU’s effect is obvious and the LRU have been widely used in practice.Disadvantage: Periodical and episodic batch operation?will cause the?LRU hit?rate decreasing sharply and the serious cache pollution?situation.The most common implementation is to use a linked list to cache the data, the detailed implementation of the algorithm as following:1 New data is inserted into the list head;2 Whenever the cache hit (i.e. the cache data is accessed), data will be moved to the list head;3 When the list is full, the tail data of the list data is discarded.2.1.3 LF