Business protocol index data collecting method and system

Abstract

The invention provides a business protocol index data collecting method and system. The method comprises the following steps: using acquired business traffic to drive state transition of a protocol stack by use of a passive protocol stack analysis technology; correctly subjecting the traffic data to a source procedure and a target procedure, and automatically bringing the service of the target procedure into a monitoring object so as to realize the automatic discovery of the monitoring object; automatically deducing the business association information between the services after correctly subjecting the traffic data to the source and target procedures; correctly recognizing an abnormal source service object according to the association between the services of the procedures when a linkage abnormality is produced, thereby helping an operation and maintenance staff to fast position the abnormality; acquiring the traffic data by use of a pure bypass way, wherein the monitoring way has no any invasive requirement on the business party; and inserting a traffic data copying module in an Linux kernel module so as to acquire the complete physical traffic at the two-layer packet level, wherein the acquired data is more complete.

Description

technical field [0001] The invention relates to a method and system for collecting service protocol index data. Background technique [0002] With the rapid development of technology and business in the Internet industry, more and more people are beginning to enjoy the convenience brought by Internet online services, and online service providers are also facing increasing pressure from business requests. In order to adapt to the increasing business pressure, various service providers have begun to migrate and evolve their server-side architectures in the direction of microservices and distribution. [0003] Although the server-side architecture can obtain horizontal scaling capabilities through distributed and micro-service methods, usually the distributed architecture will bring additional operation and maintenance complexity to the system, and the interaction between different components will become more and more complicated. And these complexities will increase with the ...

AI Technical Summary

Problems solved by technology

[0012] 3. Even if the business party embeds the business-level indicator code, it is usually difficult to obtain some relatively low-level indicator data (because the underlying details are usually encapsulated by the operating system kernel);

[0013] 4. Even if the business party embeds the business-level indicator code, the monitoring system cannot understand the specific meaning of the indicators independently provided by the business party;

[0014] 5. If the analysis of business logs (or other bypass data generated by business services) is used for indicator collection, then the collection module is difficult to be universal (because the log is a non-standardized data structure);

[0015] 6. Without intruding into the business and architecture, it is impossible to obtain related information between businesses, such as topology;

[0016] 7. The services that need to be monitored cannot be automatically discovered, and most of them need to be manually configured, which is usually very time-consuming and labor-intensive in large-scale clusters

[0018] 1. The business side needs to spend additional research and development efforts to adapt the monitoring system;

[0019] 2. For a system that has been running for a long time, intrusive monitoring solutions are usually difficult to promote;

[0020] 3. The monitoring system cannot understand the operation mode and structure of the business, and cannot make relatively intelligent judgments;

[0021] 4. Operation and maintenance personnel need to spend a lot of energy on the monitoring system for configuration work (configuration of monitoring objects, etc.)

Images