A Multidimensional Performance Diagnosis Analysis Method

Abstract

The invention discloses a multi-dimensional performance diagnosis and analysis method, which belongs to the field of performance analysis and processing of application software. The invention solves the problem of how to better analyze and process the performance of a server system. The technical solution is: a method for multi-dimensional performance diagnosis and analysis, including the following steps: the client calls the unified entry of the server, and the batch flag of the debugging mode is turned on; the application server adds interception processing, and adds a mark and time record for each request; data access Layer public interface, according to the current thread context to obtain the client's batch ID and request ID, record all SQL scripts and execution time; count resource usage and database waiting events; the client ends the debugging mode, and requests this debugging mode to the server The performance data generated during the process, combined with the performance response of the client itself, calculates and displays the time distribution and corresponding waiting conditions during the debugging request response processing process.

Description

technical field [0001] The invention relates to the field of performance analysis and processing of application software, in particular to a multi-dimensional performance diagnosis and analysis method. Background technique [0002] In the performance analysis and processing of application software, there are the following problems: [0003] 1. When the system is huge and the business is complex, the whole set of software is developed in layers, and the codes of each layer increase the log statistics response time and output, which is very inconvenient, unscientific, and the workload is heavy; [0004] 2. The tracking results of the outer interface show that the application server responds slowly, but the code I maintain calls the interfaces of other modules in many places, so I have no way to start; when communicating with colleagues in other modules, the description is complicated and communication is difficult; [0005] 3. There is an obvious delay between the response of...

AI Technical Summary

Problems solved by technology

[0003] 1. When the system is huge and the business is complicated, the entire software is developed in layers, and the codes of each layer increase the log statistics response time and output, which is very inconvenient, unscientific, and the workload is heavy;

[0004] 2. The tracking results of the outer interface show that the application server responds slowly, but the code I maintain calls the interfaces of other modules in many places, so I have no way to start; when communicating with colleagues in other modules, the description is complicated and communication is difficult;

[0005] 3. There is an obvious delay between the response of the APP Server and the DB. If you start with the code review, the workload is huge, and the subjective factors are greatly affected;

[0006] 4. It is inconvenient to track the execution records of SQL scripts in the Oracle environment. Although Sqlserver can do it, it does not know which method is called by tracking the SQL of other modules.