Cache management method, device and apparatus and computer readable storage medium

Abstract

The invention discloses a cache management method, which comprises the following steps of: distributing data in a cache space to a specified cache queue, and setting a state and index information of the data so as to index the cache queue in which the data is located and the state of the data through the index information; judging whether the space utilization rate of the cache space reaches a preset adjustment threshold value or not; if the preset adjustment threshold value is reached, adjusting a cache queue where the data is located and / or the state of the data; judging whether the space utilization rate of the cache space reaches a preset elimination threshold value or not; and if the preset elimination threshold value is reached, deleting the specified data in the cache queue. The invention can avoid cache pollution and improve the management efficiency. The invention furthermore discloses a cache management device and apparatus, and a computer readable storage medium, which all have the above technical effects.

Description

Technical field [0001] The present application relates to the field of storage technology, and in particular, to a cache management method; there is also a cache management device, device, and a computer readable storage medium. Background technique [0002] In a normal sense, the structure for coordinating the difference in data transmission speed differences for the treatment of the Northern San Sany-South speed difference is called a cache. For example, memory can be used as a cache between the hard disk and the CPU. An important indicator of cache performance is a hit rate, that is, the probability of hitting data to be accessed in the cache. The optimal scene is that all to access data occurs in the cache, which requires a better cache management algorithm to presence the client's data access behavior, so that the data is pre-judged in advance to the cache, lift the hit rate, hit rate Higher performance transitions have been made to the system. However, the cache space is li...

AI Technical Summary

Problems solved by technology

Although the efficiency of LRU is very good when there are hot data from the perspective of hit rate, sporadic and periodic batch operations will cause the LRU hit rate to drop sharply

The advantage of LRU is that it is simple to implement, but the cost is low search efficiency and cache pollution, that is, it cannot identify hot and cold data, and the management efficiency is low

Images