A caching method for capturing hot data by adaptive classification reuse distances

Abstract

The invention discloses a caching method for capturing hot data by adaptive classification reuse distances, aiming at the problem of caching strategy in computer storage field. Because most of the traditional caching strategies only consider the new progress (Recency) or frequency reflected by the data, are unable to capture the pattern characteristics reflected by the data access in a deeper level, and coupled with the algorithm itself does not have the ability to make adaptive adjustments to the captured changes in data characteristics, which may lead to the phenomenon that the data hit rateand stability of the strategy is not high. Aiming at this problem, this caching method uses the data reuse distance feature and the metadata history information replaced by the cache to mine the characteristics of the data access pattern, and processes the process of hot data identification and caching time allocation in an adaptive way, so as to improve the hit rate of the algorithm.

Description

technical field [0001] The present invention relates to the technical field of storage systems, in particular to a caching method for capturing hot data by adaptive classification reuse distance (Reuse-distance). Background technique [0002] Caching technology is not limited to high-speed caches. Since the modern computer system proposed the memory hierarchy, CPU register sets, caches, main memory, SSD and HDD storage devices, and even remote secondary storage devices such as distributed file systems can all be understood. Because it exists as a cache for larger and slower storage devices at the next layer, all storage devices themselves have cache problems. As the core part of caching technology research, the cache replacement algorithm is related to computer storage systems, operating systems, file systems, databases, web servers, and even various other application technologies such as hot and cold data identification and data compression. have a direct connection. [0...

AI Technical Summary

Problems solved by technology

However, although these excellent algorithms have many advantages, there are also problems such as not being able to capture hot data larger than the cache and cache pollution.

[0005] Therefore, aiming at the current large-scale I / O data stream environment, how to change the problem of low data hit rate caused by traditional caching technology under the condition of reducing storage system memory overhead and computing overhead, and design a novel and effective thermal identification method. It is particularly important to improve computer storage performance by data caching scheme

Images