A saas software performance fault diagnosis method based on gbdt decision tree

Abstract

In order to meet the performance maintenance requirement of SaaS software. The invention discloses a SaaS software performance fault diagnosis method based on a GBDT decision tree in combination witha performance log. A monitoring means is adopted to obtain a performance log when the software system runs. The performance fault type of the performance log is marked by combining the analysis resultof the warning log. the performance logs are processed by using a same type mean value filling method and a combined SMOTE and pre-sampling method; complete and balanced performance log data can be provided. A GBDT algorithm in a machine learning method is used for analyzing performance logs, a performance fault diagnosis model is established, the performance logs generated by a system in real time are input into the established diagnosis model, an output corresponding performance fault type is obtained, and therefore the performance state of the SaaS software in the running process is diagnosed. Besides, the real-time performance log and the diagnosis result can be saved, the diagnosis model can be updated every a period of time, the real-time performance of the diagnosis model is guaranteed, and the accuracy of the diagnosis result is further guaranteed.

Description

technical field [0001] The invention belongs to the field of software engineering, in particular to a SaaS software fault diagnosis method based on performance logs. Background technique [0002] Since the 21st century, with the vigorous development of Internet technology and the maturity of application software, an innovative software application model SaaS model has begun to emerge. SaaS is a software deployment model based on Web delivery that provides software through webservice services. Users host, provide and access the built application software through the network. SaaS has developed rapidly with its unique model, and more and more people or enterprises choose the SaaS model, so the application software under the SaaS model (hereinafter referred to as SaaS software) needs to have higher performance, reliability and availability. [0003] However, compared with the traditional model, SaaS software needs to face more challenges when running, because SaaS software wil...

AI Technical Summary

Problems solved by technology

[0003] However, compared with the traditional model, SaaS software needs to face more challenges when running, because SaaS software will cause software service quality degradation or even software performance degradation due to various factors: on the one hand, it may be caused by the software’s own architecture and It is caused by defects in code design. This problem needs to start with the software construction process and improve its architecture design. On the other hand, it may be caused by SaaS software running in a large-scale, highly complex and unpredictable dynamic cloud environment. The possible situations are as follows: (1) Insufficient resources on the virtual machine or physical node; (2) Too frequent service requests to the server; (3) Dynamic changes in the operating status of hardware resources, etc.

However, since SaaS software is mostly in a distributed cluster environment, the application software interactions between various layers are frequent, resulting in massive log data generated by each component in the system, which not only increases the difficulty of performance fault diagnosis, but also traditional fault diagnosis methods It is difficult to carry out real-time and comprehensive fault diagnosis of the system

Images