Online early warning processing method and system thereof

Abstract

The invention discloses an online early warning processing method and a system thereof. The method comprises the following steps: receiving early warning information; carrying out problem positioningaccording to the early warning information to obtain positioning information; determining whether misinformation exists or not according to the positioning information, and if not, selecting a processing plan according to the positioning information; performing the processing plan.

Description

technical field [0001] This manual relates to the field of computer technology, in particular to online risk control based on the Internet. Background technique [0002] Internet-based large-scale online service systems have a large number of users. For example, the number of Alipay users currently exceeds 800 million. User behavior, external changes, internal changes, and daily operation and maintenance activities may cause online abnormalities. These abnormalities may directly lead to economic loss. How to stop bleeding and stop loss online immediately after receiving the warning, especially how to stop loss at the minute level, has always been a challenging problem. [0003] At present, there are two ways to deal with early warning. One is manual method. After online monitoring and early warning, it is necessary to manually judge whether the early warning is effective. It often needs to combine expert experience, online operation log, database record query and other mean...

AI Technical Summary

Problems solved by technology

[0003] At present, there are two ways to deal with early warning. One is manual method. After online monitoring and early warning, it is necessary to manually judge whether the early warning is effective. It often needs to combine expert experience, online operation log, database record query and other means to make a comprehensive judgment, and make a further decision whether to Emergency response is required. The emergency process often requires steps such as pushing the switch, refreshing the cache, and correcting the data. The entire process is time-consuming and relies heavily on human experience and operations, resulting in poor flexibility.

[0004] The other is the automatic method, which is to simply associate the early warning with the treatment plan. Once the early warning is monitored, the associated treatment plan will be automatically executed. Because the monitored early warning often has false positives (or noise), it will cause problems in the implementation of the plan. Equal-proportional noise can easily cause unexpected impacts on the online system, making it difficult to promote on a large scale

Images