Cache management method and system for block-level data

Abstract

The invention discloses a cache management method and a cache management system for block-level data, wherein the method comprises the following steps: adding an SSD (solid state disk) as a second-level cache between a first-level cache and a disk array of a storage server; when a client side sends a writing request, writing a data block into the second-level cache having a free space if the judgment is performed that the data block required to be written does not exist in a client side cache, a first-level cache of a home terminal and a second-level cache of the home terminal; when the client side sends a reading request, reading the data block from the disk array to the second-level cache if the judgment is performed that the data block required to be read does not exist in the client side cache, the first-level cache of the home terminal and the second-level cache of the home terminal; and floating the data block in the second-level cache upwards to the client side cache if the judgment is performed that the data block required to be read exists in the second-level cache. According to the cache management method and the cache management system, the redundancy of the data block of the client side in the system is reduced.

Description

technical field [0001] The invention relates to the field of computer application technology, in particular to a buffer management method and system based on block-level data. Background technique [0002] Considering the current situation that the storage server based on the solaris operating system regards the memory as a cache, the capacity of the cache is limited to a certain extent. At this stage, the caching mechanism on the storage server and the exclusive mechanism strategy of data blocks on all levels of caching have the following shortcomings: [0003] (1) Most of the storage servers based on the solaris system use the memory as a cache, which is why the memory usage of the storage server system is relatively high. But in this way, other problems have been brought about one after another: first, the price of memory is relatively high, which will increase the cost investment of customers; second, it will reduce the response speed of the storage server system to use...

AI Technical Summary

Problems solved by technology

But in this way, other problems have been brought one after another: first, the price of memory is relatively high, which will increase the cost investment of customers; second, it will reduce the response speed of the storage server system to user processing events

[0004] (2) In the case of coexistence of multi-level caches in the system, the current caching strategy cannot efficiently improve the ability of the storage server to read data. The multi-level caching mechanism will cause a data block to be stored in multiple cache levels, that is, the data Blocks exist in multiple copies throughout the cache structure

The traditional exclusive cache strategy needs to sacrifice the performance of the server's local cache because the exclusive cache mechanism cannot be well implemented without modifying the client software and I / O interface; or in order to ensure the high performance of the storage server's local cache , it is necessary to modify the client software and I / O interface; the two cannot be balanced, resulting in inefficient data replacement strategies

Images