API monitoring method, system and equipment based on relative anomaly thought and storage medium

Abstract

The invention relates to an API monitoring method, system and equipment based on a relative anomaly thought and a storage medium. In order to solve the problem that traditional API monitoring software cannot meet large-scale, high-complexity and periodic system monitoring requirements, the invention provides a scheme that fully considers the complexity, periodic fluctuation and the like of all interfaces; except for traditional absolute information such as an API list with the highest transaction volume, an API list with the lowest success rate and an API list with the highest average time consumption, the relative anomaly thought is innovatively introduced into monitoring; according to the invention, relative information such as a newly-added calling API list, an API list with transaction volume abnormal change, an API list with success rate abnormal change, an API list with average time consumption abnormal change and the like can be quickly, scientifically and effectively displayed; various personalized custom control parameter combinations are supported, and results can be analyzed more effectively with less human input and higher monitoring efficiency; the invention plays an important role in scenes such as API technical verification after production, production problem real-time / afterward positioning and the like.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to an API monitoring method, system, device and storage medium based on the idea of ​​relative abnormality. Background technique [0002] API is a product of the rise of the Internet and cloud computing, which makes the interaction between software applications possible. Almost every code-level interaction process collects data or triggers certain key processes by calling APIs. Software's dependence on API is not only manifested in the number of APIs and calls, but also in the role positioning of APIs. If there is a problem with the front-end, it will generally only affect the display of some functions, and if there is a problem with the back-end API, it will often directly It affects the correctness of business logic, whether the service can respond in time, and even the availability of the entire software. [0003] With the explosive growth of mobile applications, more ...

AI Technical Summary

Problems solved by technology

The problem brought about by the former is that the success rate / time-consuming ranking loses its due meaning. A complex interface A that takes an average of 1000ms, and a simple interface B that takes an average of 100ms, regardless of the particularity of each interface, Then even if the abnormality of interface B takes 200ms, interface A will always be in front of interface B.

The problem caused by the latter is that the normal fluctuation factors of the API are not excluded. For example, individual transactions are concentrated in a specific time period every Friday, then the interface involved in the transaction will have a significant increase in transaction volume during this period As well as the increase in time consumption, it will return to the normal level after the transaction time period. If the API operation is monitored during this time period, the interface will have obvious abnormal performance degradation

The ARMS-API monitoring solution cannot handle the above two special cases, resulting in a large gap between the given ranking information and the 'true anomaly'. For a large-scale complex system with periodic fluctuations, its positioning in the abnormal interface, etc. The role that can be played in the scene is very limited

Images