Threshold-free detection method and device for cloud platform

Abstract

The invention discloses a threshold-free detection method for a cloud platform, and the method comprises the steps: obtaining to-be-detected index data, and establishing models through a plurality ofabnormity detection algorithms; performing unsupervised abnormity detection on the to-be-detected index data according to the model; when any model obtains abnormal index data from the to-be-detectedindex data through unsupervised anomaly detection, judging whether a supervised model trained through corresponding historical index data exists according to the abnormal index data; when it is judgedthat no supervision model exists, all results of unsupervised abnormity detection are integrated to judge whether the to-be-detected index data are abnormal; when it is judged that the supervised model exists, performing supervised abnormity detection on the abnormal index data according to the supervised model, and judging whether the to-be-detected index data are abnormal according to a supervised abnormity detection result. The invention further discloses a corresponding device. According to the method, the joint detection method of the unsupervised model and the supervised model is realized, and the accuracy of the threshold-free detection result is improved.

Description

technical field [0001] The present invention relates to the field of computer technology, and more specifically, to a method and device for non-threshold detection on a cloud platform. Background technique [0002] OpenStack is an open source cloud computing management platform project, which is a combination of a series of software open source projects. The goal of the project is to provide a cloud computing management platform that is simple to implement, scalable, rich, and standardized. [0003] As OpenStack becomes more and more mature, the scale of OpenStack-based cloud platforms continues to expand, from the initial few or a dozen to hundreds or even thousands. As a result, the types and volume of monitoring data are increasing, and the alarm rules are becoming more and more complex. [0004] The performance data of the cloud platform (memory usage rate, CPU (Central Processing Unit, central processing unit) system (system) usage rate, network inflow speed, etc.) is...

AI Technical Summary

Problems solved by technology

However, in the actual application process, it is found that the threshold setting is too dependent on experience, the threshold is too high, there are many missing alarms, and it is difficult to find quality risks, and the threshold is too low, too many alarms cause an alarm storm, which interferes with the judgment of business operation and maintenance personnel; In the case of resource jitter, threshold alarms cannot be detected, and false positives will occur.

Images