A method and device for business system performance optimization

Abstract

The invention discloses a method and a device for optimizing the performance of a service system. The method includes: judging whether the first kernel function called during the first time period of the first loop cycle belongs to the most frequently called in the first time period among the multiple cycle cycles before the first loop cycle The first n kernel functions; if the first kernel function belongs to the top n kernel functions that are called the most times in the first time period in the multiple cycle cycles before the first loop cycle, then in the kernel function access history list , performing a weighting operation on the number of calls of the first kernel function; wherein, the kernel function access history list is used to determine the kernel function to be loaded into the cache of the CPU. By weighting the number of calls of the first kernel function in the kernel function access history list, the cache hit rate of the CPU of the business system is improved, the efficiency of reading data of the business system is optimized, and the performance is improved. optimization.

Description

technical field [0001] The invention relates to a data storage technology of a business system, in particular to a method and a device for performance optimization of a business system. Background technique [0002] When the operating system of the business system is running, it will use many storage areas, such as hard disk, U disk, memory, and CPU cache (cache, also called high-speed cache); among them, the reading speed of CPU cache is faster than that of non-cached storage. Areas (for example, hard disk, U disk, memory, etc.) can be read much faster. Therefore, when the operating system needs to read data, it will first search for the required data from the CPU cache, and if it finds it, it will directly read it. Read the data, and if not found, look it up from memory or other storage areas. In the cache mechanism, there will be a phenomenon of reading infrequently used data into the cache, which is called cache pollution. Because infrequently used data occupies the ca...

AI Technical Summary

Problems solved by technology

When the LINUX operating system executes an instruction that is executed periodically, it is possible that when the instruction that is executed periodically is just loaded into the cache of the CPU, this instruction is rarely called in the following period of time, so that the instruction that is executed periodically The CPU cache where the instruction is located is emptied, and this emptied CPU cache is loaded with other instructions, which causes the periodically executed instruction to be cleared before being used after being loaded into the CPU cache, which will cause As a result, the hit rate of the CPU cache will drop sharply, and the cache pollution is serious.

Images