• Least Recently Used (LRU)
• Clock Page Replacement
• 2Q
• Adaptive Replacement Cache (ARC)

• This repository aims to analyse different page replacement algorithms.
• This repository contains implementations of the said page replacement algorithms, created in accordance with the relevant research papers.

• Paging Experiment with the Multics System
• Least Recently Used
• 2Q - A Low Overhead High Performance Buffer Management Replacement Algorithm
• A Cache Algorithm for Disk-backed Databases
• ARC: A Self-tuning, Low Overhead Replacement Cache
• Google benchmark

• Clone this Repository.

  $ git clone [email protected]:Inamul07/Cache_Analysis.git
  $ cd Cache_Analysis

• Edit the main.c file to include the required replacement algorithm.
• Then run this command in terminal,

  $ make
  

• cache_init(cacheType cache_name, int capacity)
  • Initializes the respective cache with given capacity.
  • returns generic_cache*.
• cache_access(generic_cache* cache, int page)
  • Performs the relevant cache operation.
  • Keeps an internal count of number of hits and misses.
  • returns void.
• cache_analysis(generic_cache* cache)
  • Prints the buffer, total references, number of hits and misses and hit ratio.
  • returns void.
• cache_print_buffer(generic_cache* cache)
  • Prints the buffer at that current state.
  • returns void.
• cache_put_array(generic_cache* cache, int pages[], int size)
  • Performs the relevant cache operation for each data item from pages in a linear fashion.
  • returns void.
• cache_destroy(generic_cache* cache)
  • free the memory allocated by the cache.
  • returns void.

For Cache Initialisation, Pass the appropriate cache_name to the cache_init() method. ( Refer Table for appropriate cache_name )
Example:

• For Initialising LRU with a capacity of 3, cache_init(LRU, 3);
• For Initialising TWO_QUEUE with a capacity of 5, cache_init(TWO_QUEUE, 5);.

• The data folder, at the root, contains test data generated by DataGenerator.java.
• We use Google benchmark to benchmark the Cache Replacement Algorithms.
• To run bench mark, execute the following command

  $ make benchmark

• This prints the time taken by the cache algorithms along with their respective Hit Ratio.