Operation system performance analysis method and device

Abstract

The invention provides an operation system performance analysis method and device. The method comprises the following steps: when a preset event in an operation system enables a performance counter in a central processing unit (CPU) to overflow, triggering to generate non maskable interrupt (NMI); during the interruption period of the NMI, sampling progress parameter of the CPU to obtain samplingdata; according to a process identification in the sampling data, acquiring command information of the current progress corresponding to the process identification; according to the command information, carrying out classified statistic on the sampling data to obtain statistic information; and according to the statistic information, analyzing the performance of the operation system. In the invention, the progress parameter is sampled when the preset event happens to obtain the statistical information for analysis, so that factors causing performance bottlenecks can be known under the condition of no error log, thereby improving the quality and efficiency of the operation system performance analysis.

Description

technical field [0001] The invention relates to an operating system performance analysis method and device, belonging to the technical field of computers. Background technique [0002] With the continuous development of the Internet, the scale of network applications is gradually expanding. A complete network application system may consist of dozens of subsystems, and a complex subsystem may consist of millions of lines of code. The expansion of the application scale and the increase of the code volume make the operating system debugging more and more difficult. In the development process, there are many ways to debug errors, but in the production environment, many unknown problems and emergencies will be encountered. In addition to conventional methods such as troubleshooting based on logs, there are generally few other debugging methods, especially when When the operating system has a performance bottleneck and there is no error log, it is difficult to find out the probl...

AI Technical Summary

Problems solved by technology

In the development process, there are many ways to debug errors, but in the production environment, many unknown problems and emergencies will be encountered. In addition to conventional methods such as troubleshooting based on logs, there are generally few other debugging methods, especially when It is difficult to find out the problem when the operating system has a performance bottleneck and there is no error log

Images