CN-121996501-A - Automatic performance monitoring method and device for vehicle-mounted multimedia application and storage medium

CN121996501ACN 121996501 ACN121996501 ACN 121996501ACN-121996501-A

Abstract

The application relates to the field of software performance monitoring, and provides an automatic performance monitoring method, device and storage medium for vehicle-mounted multimedia application. The method comprises the steps of collecting multi-dimensional performance parameters of a target application in real time when the target application runs, judging whether preset trigger conditions are met based on the multi-dimensional performance parameters of the target application in the running, automatically starting a plurality of data collection channels in parallel to obtain multi-mode data related to performance events when the preset trigger conditions are met, carrying out correlation analysis based on uniform time stamps on the collected multi-mode data and the multi-dimensional performance parameters to construct a complete data evidence chain of the performance events, and automatically generating a performance analysis report containing performance bottleneck positioning information and optimization suggestions based on performance event causes and influence ranges determined by the correlation analysis. The technical scheme of the application realizes accurate, efficient positioning and automatic diagnosis of the vehicle-mounted application performance problem.

Inventors

LI MEIJIAO
Liu Sande

Assignees

北京雷石天地电子技术股份有限公司

Dates

Publication Date: 20260508
Application Date: 20251226

Claims (10)

1. An automated performance monitoring method for a vehicle-mounted multimedia application, the method comprising: Collecting multidimensional performance parameters of a target application in real time when the target application runs; judging whether a preset triggering condition is met or not based on the multi-dimensional performance parameters; when the triggering condition is met, automatically starting a plurality of data acquisition channels in parallel to acquire multi-mode data related to the performance event; Performing correlation analysis based on uniform time stamps on the acquired multi-modal data and the multi-dimensional performance parameters to construct a complete data evidence chain of the performance event; And automatically generating a performance analysis report containing performance bottleneck positioning information and optimization suggestions based on the performance event causes and the influence ranges determined by the relevance analysis.
2. The automated performance monitoring method of an on-vehicle multimedia application according to claim 1, wherein the automatically starting a plurality of data acquisition channels in parallel to obtain multi-modal data related to a performance event comprises: starting a first channel, and recording snapshot data of a system level performance index at a preset frequency; starting a second channel, recording video of the operation interface of the target application, and generating an interface operation video stream; Starting a third channel, capturing and storing the log information of the target application and the bottom layer of the system, and forming a log file; And starting a fourth channel, and continuously monitoring and recording a resource occupation state change sequence of each thread in the target application.
3. The method for monitoring the automated performance of the vehicle-mounted multimedia application according to claim 2, wherein the performing the correlation analysis based on the uniform time stamp on the collected multi-modal data comprises: marking uniform high-precision time stamps on the multi-mode data and the multi-dimensional performance parameters; And based on the unified high-precision time stamp, carrying out transverse association comparison on user operation events in the interface operation video stream, error records in a log file, abnormal peaks in multi-dimensional performance parameters, abnormal data in snapshot data of a system-level performance index and blocking events in a resource occupation state change sequence so as to reconstruct a complete time line of performance event occurrence.
4. The automated performance monitoring method of an in-vehicle multimedia application of claim 3, wherein the correlation analysis further comprises: Identifying a resource use abnormal mode of a system layer based on the historical sequence data of the multi-dimensional performance parameters, wherein the resource use abnormal mode comprises a periodic high-load mode, a memory leakage trend or an interface rendering katon mode; And for the identified abnormal mode of resource use, correlating and analyzing the thread state change sequence data in the same time period to locate a thread level behavior mode which leads to the abnormal mode of resource use, wherein the thread level behavior mode comprises an execution track of a high-consumption thread, a thread competition hot spot or frequent thread creation/destruction operation.
5. The method for automated performance monitoring of an in-vehicle multimedia application according to claim 4, wherein locating thread-level behavior patterns that result in abnormal patterns of resource usage comprises: aiming at the interface rendering jamming mode, associating the time point of jamming with the time point of the CPU occupation peak value of a specific rendering thread and the time point of occurrence of graphic library error information recorded in a log file so as to locate a root thread or a code module which causes jamming; and aiming at the memory leakage trend, correlating the time period of continuous memory growth with the memory allocation operation record of the service thread which does not normally release resources, so as to locate the code path with suspected memory leakage.
6. The method for monitoring the automatic performance of the vehicle-mounted multimedia application according to claim 1, wherein the method further comprises periodically sampling to obtain CPU time slice occupation proportion, memory use condition and state transition information of each thread in the target application by adopting a low-overhead sampling mode; the low-overhead sampling mode is: Acquiring thread state data in a sampling mode, wherein in the sampling process, an intelligent filtering strategy is applied to preferentially acquire the thread data meeting the preset activity or resource consumption condition; The obtained thread state data is cached in the memory, and when the cached data quantity reaches a batch threshold value, the persistent storage operation is executed.
7. The automated performance monitoring method of an in-vehicle multimedia application according to claim 1, wherein prior to automatically generating a performance analysis report based on the determined performance event causes and impact ranges of the relevance analysis, the method further comprises: Automatically identifying a main process of the target application and a service process with a dependence or communication relationship with the main process, and constructing a process relationship map; And based on the process relation map, performing collaborative monitoring and association analysis on the performance data of the main process and the service process, and using an analysis result to determine the cause and the influence range of the performance event.
8. An automated performance monitoring apparatus for an in-vehicle multimedia application, the apparatus comprising: The acquisition module is used for acquiring multidimensional performance parameters of the target application in real time when the target application runs; The judging module is used for judging whether a preset triggering condition is met or not based on the multi-dimensional performance parameters; the acquisition module is used for automatically starting a plurality of data acquisition channels in parallel when the triggering condition is met so as to acquire multi-mode data related to the performance event; the analysis module is used for carrying out correlation analysis based on uniform time stamps on the acquired multi-mode data and the multi-dimensional performance parameters so as to construct a complete data evidence chain of the performance event; And the generation module is used for automatically generating a performance analysis report containing performance bottleneck positioning information and optimization suggestions based on the performance event cause and the influence range determined by the relevance analysis.
9. An apparatus comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when the computer program is executed by the processor.
10. A storage medium storing a computer program which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

Description

Automatic performance monitoring method and device for vehicle-mounted multimedia application and storage medium Technical Field The present application relates to the field of software performance monitoring, and in particular, to a method, an apparatus, and a storage medium for monitoring an automated performance of a vehicle-mounted multimedia application. Background With the rapid development of intelligent and networking of automobiles, the vehicle-mounted intelligent cabin has become an important mobile interaction space, and the integrated multimedia entertainment applications (such as streaming media, KTV, games, and the like) are increasingly complex, so that the requirements on functions and performances are continuously increased. These applications typically involve real-time audio and video processing, graphics rendering, multi-threaded scheduling, and deep interaction with the underlying system, with stable, smooth operation directly related to user experience and driving safety. Currently, for performance monitoring of such complex applications, conventional performance analysis tools and methods are commonly employed within the industry. The methods may include deploying an independent performance counter to collect indicators such as CPU and memory, or recording an operation state by means of a log system, or manually starting a screen by a tester to record and grab a log after a suspected problem occurs, and performing post analysis. These techniques can help to locate some of the performance problems to some extent. However, the applicant has found that the above prior art solutions have significant drawbacks when faced with the application of high performance, high real-time requirements in a vehicle cabin environment. Firstly, the existing scheme relies on manual observation or simple threshold value alarming, is difficult to automatically and instantly trigger omnibearing field data acquisition at the moment of performance abnormality occurrence, so that precious first field information (such as the instant state of a specific thread and the corresponding relation between user operation and interface response) is easy to lose, secondly, acquired data such as performance indexes, system logs, interface display and the like are often isolated and fragmented, and lack of accurate time synchronization and effective association, so that an analyst is difficult to quickly construct a complete evidence chain reflecting the overall appearance of a problem from multidimensional data, and the problem positioning efficiency is low and root cause analysis is difficult. Disclosure of Invention The application provides an automatic performance monitoring method, an automatic performance monitoring device and a storage medium for vehicle-mounted multimedia application, which realize accurate, efficient positioning and automatic diagnosis of vehicle-mounted application performance problems by automatically triggering multidimensional data acquisition and association analysis. In one aspect, the present application provides a method for monitoring the automation performance of a vehicle-mounted multimedia application, the method comprising: Collecting multidimensional performance parameters of a target application in real time when the target application runs; judging whether a preset triggering condition is met or not based on the multi-dimensional performance parameters; when the triggering condition is met, automatically starting a plurality of data acquisition channels in parallel to acquire multi-mode data related to the performance event; Performing correlation analysis based on uniform time stamps on the acquired multi-modal data and the multi-dimensional performance parameters to construct a complete data evidence chain of the performance event; And automatically generating a performance analysis report containing performance bottleneck positioning information and optimization suggestions based on the performance event causes and the influence ranges determined by the relevance analysis. In another aspect, the present application provides an automated performance monitoring apparatus for a vehicle-mounted multimedia application, the apparatus comprising: The acquisition module is used for acquiring multidimensional performance parameters of the target application in real time when the target application runs; The judging module is used for judging whether a preset triggering condition is met or not based on the multi-dimensional performance parameters; the acquisition module is used for automatically starting a plurality of data acquisition channels in parallel when the triggering condition is met so as to acquire multi-mode data related to the performance event; the analysis module is used for carrying out correlation analysis based on uniform time stamps on the acquired multi-mode data and the multi-dimensional performance parameters so as to construct a complete data evidence chain of the pe