A browser-based multi-dimensional application access and performance data collection and analysis system, method, device and medium

Abstract

The application relates to the technical field of browser operation, and discloses a multi-dimensional application access and performance data collection and analysis system, method, equipment and medium based on a browser. The system injects a lightweight monitoring script when a page is loaded, dynamically collects user access behaviors, page performance indexes and network error data, carries out multi-dimensional aggregation on the above data based on preset or self-defined time dimensions, and can generate visual charts such as line charts, column charts and pie charts. Whether an optimization strategy is triggered can be automatically judged according to a performance baseline and an error rate threshold, and a structured suggestion can be pushed through linkage of an alarm module, so that data support is provided for improvement of web application stability and user experience satisfaction.

Description

Technical Field

[0001] This invention relates to the field of browser operation technology, and in particular to a browser-based multidimensional application access and performance data acquisition and analysis system, method, device and medium. Background Technology

[0002] Current web applications generally adopt a front-end and back-end separation architecture. This approach results in a large amount of business logic running on the browser side, which makes it impossible to accurately reflect the actual user experience by relying solely on server logs.

[0003] To address this deficiency, some researchers in the industry have begun to introduce front-end performance monitoring solutions. However, existing solutions are typically limited in functionality; some systems only collect page load times, while others only capture JavaScript errors. They lack the ability to collect and correlate user access behavior, performance metrics, and network anomalies in a unified manner.

[0004] Due to the fragmented data dimensions, it is difficult for operations and maintenance personnel to determine whether the performance degradation is caused by a specific error, let alone identify how the high error rate is caused.

[0005] In addition, most existing tools do not support flexible aggregation by day, month or custom time range, which limits trend analysis. At the same time, these tools have weak visualization capabilities and cannot intuitively reveal the relationship between traffic changes and performance degradation through charts. Summary of the Invention

[0006] In view of the aforementioned existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a browser-based multidimensional application access and performance data acquisition and analysis system, method, device and medium, which can solve the problems of limited functionality and low fault location efficiency in existing front-end performance monitoring solutions.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a browser-based multidimensional application access and performance data acquisition and analysis system, the system comprising a data acquisition module, a first statistics module, a first visualization module, a second statistics module, a second visualization module, and a third determination module, wherein: The data acquisition module is used to dynamically collect user access behavior data, page performance index data, and network error data of the target application through browser-side scripts; The user access behavior data includes unique visitors (UV) and page views (PV); The page performance metrics data include DNS resolution time, SSL connection time, and first byte time (TTFB). The network error data includes HTTP status codes and JavaScript error messages; The first statistics module is used to perform multi-dimensional aggregation based on the user access behavior data and the page performance index data, combined with a preset time dimension or a custom time range, to obtain application access and performance statistics data. The first visualization module is used to generate and display multi-dimensional visualization charts based on the application access and performance statistics. The multi-dimensional visualization charts include line charts for presenting access trends and bar charts for presenting the distribution of performance bottlenecks.

[0009] As a preferred embodiment of the browser-based multidimensional application access and performance data acquisition and analysis system described in this invention, it further includes: The second statistics module is used to classify and statistically analyze the network error data to obtain error analysis statistics. The error analysis statistics include the total number of errors, the percentage of affected users, the top 5 distribution of abnormal status codes, and the distribution of client environments of users who reported errors. The second visualization module is used to generate and display anomaly distribution visualization charts based on the error analysis statistics. The anomaly distribution visualization charts include a bar chart for presenting the distribution of abnormal status codes and a pie chart for presenting the distribution of the user client environment of the error. The third determining module is used to determine the target optimization strategy based on the application access and performance statistics and the error analysis statistics, so as to improve the stability and user experience of the target application.

[0010] As a preferred embodiment of the browser-based multidimensional application access and performance data collection and analysis system of the present invention, wherein: in the aspect of dynamically collecting user access behavior data, page performance index data, and network error data of the target application through browser-side scripts, the data collection module is specifically used for: Inject a performance monitoring script when the target application's page loads; The performance monitoring script listens for page lifecycle events and captures page performance metrics data. The performance monitoring script listens for error events in the page's global object, captures JavaScript error information, and records the corresponding device information and error stack. The performance monitoring script records the user's page navigation behavior and generates access detail logs containing the destination URL, visitor IP address, and MAC address. The page performance metrics data, the JavaScript error information, and the access details log are reported to the server as raw data.

[0011] As a preferred embodiment of the browser-based multidimensional application access and performance data collection and analysis system of the present invention, wherein: in the aspect of combining a preset time dimension or a custom time range for multidimensional aggregation to obtain application access and performance statistics, the first statistical module is specifically used for: Obtain the time dimension selected by the user, which includes any one of days, months, and years, or a start and end time range specified by the user; Based on the time dimension, the user access behavior data is aggregated by UV and PV to obtain access statistics. Based on the time dimension, the page performance index data is calculated by quantiles according to DNS resolution time, SSL connection time and TTFB to obtain performance index statistics. The access statistics are correlated with the performance metrics statistics to form the application access and performance statistics.

[0012] As a preferred embodiment of the browser-based multidimensional application access and performance data collection and analysis system of the present invention, wherein: in the process of classifying and statistically analyzing the network error data to obtain error analysis statistics, the second statistical module is specifically used for: The HTTP status codes are filtered to determine that requests with a status code greater than or equal to 400 are abnormal requests. The total number of errors is obtained by counting the total number of abnormal requests; Based on the user identifier associated with the abnormal request, the proportion of users who have made at least one abnormal request to the total number of active users is calculated to obtain the proportion of affected users. The HTTP status codes of the abnormal requests are sorted by frequency, and the top five status codes and their frequencies are taken to obtain the TOP5 distribution of the abnormal status codes. Extract the JavaScript error information and the client environment information contained in the abnormal request. The client environment information includes the operating system type and browser version. Count the number of errors under each environment combination to obtain the client environment distribution of the user who reported the error.

[0013] As a preferred embodiment of the browser-based multidimensional application access and performance data collection and analysis system of the present invention, wherein: in determining the target optimization strategy based on the application access and performance statistics and the error analysis statistics, the third determining module is specifically used for: Obtain the preset performance baseline threshold and error rate alarm threshold; The performance metric statistics are compared with the performance baseline threshold. If at least one performance metric exceeds the performance baseline threshold, a performance optimization strategy is triggered. The percentage of affected users is compared with the error rate alarm threshold. If the percentage of affected users exceeds the error rate alarm threshold, a troubleshooting strategy is triggered. If both the performance optimization strategy and the fault diagnosis strategy are triggered simultaneously, a comprehensive optimization suggestion is generated as the target optimization strategy.

[0014] As a preferred embodiment of the browser-based multidimensional application access and performance data acquisition and analysis system of the present invention, the system further includes an alarm linkage module, which is used for: Once the third determining module determines the target optimization strategy, it generates corresponding alarm information based on the type of the target optimization strategy. The alarm information is pushed to a preset operation and maintenance terminal, and the multi-dimensional visualization chart and the abnormal distribution visualization chart can be exported as attachments along with the alarm information.

[0015] Secondly, this invention provides a browser-based method for collecting and analyzing multi-dimensional application access and performance data, including: Acquire user access behavior data, page performance index data, and network error data of the target application dynamically collected via browser-side scripts; The user access behavior data includes unique visitors (UV) and page views (PV); The page performance metrics data include DNS resolution time, SSL connection time, and first byte time (TTFB). The network error data includes HTTP status codes and JavaScript error messages; Based on the user access behavior data and the page performance index data, multi-dimensional aggregation is performed in combination with a preset time dimension or a custom time range to obtain application access and performance statistics. Based on the application access and performance statistics, multi-dimensional visualization charts are generated and displayed, including line charts to present access trends and bar charts to present the distribution of performance bottlenecks. The network error data is classified and statistically analyzed to obtain error analysis statistics; the error analysis statistics include the total number of errors, the percentage of affected users, the distribution of the top 5 abnormal status codes, and the distribution of the client environment of the users who reported errors. An anomaly distribution visualization chart is generated and displayed based on the error analysis statistics. The anomaly distribution visualization chart includes a bar chart to present the distribution of anomaly status codes and a pie chart to present the distribution of the user client environment of the user who reported the error. Based on the application access and performance statistics and the error analysis statistics, a target optimization strategy is determined to improve the stability and user experience of the target application.

[0016] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0017] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0018] Compared to existing technologies, the beneficial effects of this invention are that it proposes a browser-based multi-dimensional application access and performance data collection and analysis system. This system dynamically collects user access behavior, page performance metrics, and network error data by injecting a lightweight monitoring script during page loading. Based on preset or custom time dimensions, it performs multi-dimensional aggregation of the above data, generating visual charts such as line charts, bar charts, and pie charts. It can also automatically determine whether to trigger optimization strategies based on performance baselines and error rate thresholds, and link with the alarm module to push structured suggestions, thereby providing data support for improving web application stability and user experience satisfaction. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the system structure of a browser-based multidimensional application access and performance data acquisition and analysis system, as provided in one embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the injection and activation process of a browser-based multidimensional application access and performance data acquisition and analysis system, provided as an embodiment of the present invention.

[0022] Figure 3This is a schematic diagram illustrating the steps of a browser-based multidimensional application access and performance data collection and analysis system provided in one embodiment of the present invention, in which a performance monitoring script captures page performance index data by listening to page lifecycle events.

[0023] Figure 4 This is a schematic diagram illustrating the steps of a performance monitoring script in a browser-based multidimensional application access and performance data acquisition and analysis system, provided as an embodiment of the present invention, to capture JavaScript error information by listening to error events of the page's global object and recording the corresponding device information and error stack.

[0024] Figure 5 This invention provides a schematic diagram illustrating the steps of a performance monitoring script for a browser-based multidimensional application access and performance data collection and analysis system. The script hijacks the browser's native navigation mechanism to record user page navigation behavior and generates access detail logs containing the destination URL, visitor IP address, and MAC address.

[0025] Figure 6 This is a schematic diagram illustrating the steps of a browser-based multidimensional application access and performance data collection and analysis system provided in one embodiment of the present invention. The first statistical module aggregates user access behavior data according to the time dimension selected by the user, based on the number of unique visitors (UV) and the number of page views (PV), to obtain access statistics.

[0026] Figure 7 This is a schematic diagram illustrating the steps of a browser-based multidimensional application access and performance data collection and analysis system provided in one embodiment of the present invention. The first statistical module calculates the quantiles of page performance index data according to the user-selected time dimension, based on DNS resolution time, SSL connection time, and first byte time (TTFB), to obtain performance index statistics.

[0027] Figure 8 This is a schematic diagram illustrating the steps of a browser-based multidimensional application access and performance data collection and analysis system, provided in one embodiment of the present invention, to associate access volume statistics data with performance indicator statistics data to form application access and performance statistics data.

[0028] Figure 9 This is a schematic diagram illustrating the steps of a third determining module in a browser-based multidimensional application access and performance data acquisition and analysis system to obtain preset performance baseline thresholds and error rate alarm thresholds, as provided in an embodiment of the present invention.

[0029] Figure 10 This is a schematic diagram illustrating the steps of triggering a performance optimization strategy in a browser-based multidimensional application access and performance data acquisition and analysis system, as provided in one embodiment of the present invention.

[0030] Figure 11 This is a schematic diagram illustrating the steps of a browser-based multidimensional application access and performance data acquisition and analysis system for triggering fault diagnosis strategies, as provided in an embodiment of the present invention.

[0031] Figure 12 This is a schematic diagram illustrating the steps of a third determining module in a browser-based multidimensional application access and performance data acquisition and analysis system, as provided in an embodiment of the present invention, to generate a comprehensive optimization suggestion as the target optimization strategy when simultaneously triggering performance optimization strategy and fault diagnosis strategy.

[0032] Figure 13 This is an internal structure diagram of an electronic device for a browser-based multidimensional application access and performance data acquisition and analysis system, as provided in one embodiment of the present invention. Detailed Implementation

[0033] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0034] Example 1, referring to Figures 1-12 This is the first embodiment of the present invention, which provides a browser-based multidimensional application access and performance data collection and analysis system, comprising: Figure 1 This diagram illustrates the system architecture of a browser-based multidimensional application access and performance data acquisition and analysis system. The system includes a data acquisition module, a first statistics module, a first visualization module, a second statistics module, a second visualization module, and a third determination module, wherein: The data acquisition module is used to dynamically collect user access behavior data, page performance index data, and network error data of the target application through browser-side scripts; User access behavior data includes unique visitors (UV) and page views (PV); Page performance metrics include DNS resolution time, SSL connection time, and first byte time (TTFB). Network error data includes HTTP status codes and JavaScript error messages; The first statistics module is used to aggregate application access and performance statistics based on user access behavior data and page performance index data, combined with preset time dimensions or custom time ranges, to obtain application access and performance statistics. In some embodiments, the first visualization module of the present invention is used to generate and display multi-dimensional visualization charts based on application access and performance statistics. The multi-dimensional visualization charts include line charts for presenting access trends and bar charts for presenting the distribution of performance bottlenecks.

[0035] In some embodiments, the system designed according to the present invention further includes: The second statistics module is used to classify and statistically analyze network error data to obtain error analysis statistics. Error analysis statistics include the total number of errors, the percentage of affected users, the top 5 distribution of abnormal status codes, and the distribution of client environments of users who reported errors; The second visualization module is used to generate and display anomaly distribution visualization charts based on error analysis statistics. The anomaly distribution visualization charts include bar charts to present the distribution of abnormal status codes and pie charts to present the distribution of error user client environments. In some embodiments, the third determining module in this invention is used to determine a target optimization strategy based on application access and performance statistics and error analysis statistics, so as to improve the stability of the target application and the user experience.

[0036] In some embodiments, regarding the dynamic collection of user access behavior data, page performance index data, and network error data of a target application via browser-side scripts, the data collection module is specifically used for: Inject a performance monitoring script when the target application's page loads; In some embodiments, the performance monitoring script is deployed during the initial loading of the target application page via static embedding in the HTML document header or dynamic injection on the server side. The injection and activation process is as follows: Figure 2 As shown, the specific steps include: Step 1.1: At the beginning of the HTML document parsing stage of the target application, insert a script tag in the document.head node to reference an external performance monitoring script, or have the server directly inject the script content inline when generating the HTML response; Step 1.2: After the browser parses the script tag, it immediately executes the script code, establishes an independent execution context, and binds the monitoring logic to the window global object scope; Step 1.3: Determine if the current page has completed the domLoading state. If not, proceed to step 1.4; otherwise, proceed to step 1.7. Step 1.4: Register listeners for key event points of the PerformanceTiming interface, including navigationStart, domainLookupStart, secureConnectionStart, responseStart, and domComplete; Step 1.5: Override the window.history.pushState and replaceState methods to capture the destination URL when the front-end route jump is triggered, and synchronously record the visitor's IP address and MAC address to generate a structured access details log; Step 1.6: Bind the window.onerror and unhandledrejection event handlers to capture JavaScript synchronous exceptions and unhandled Promise rejections, and extract the operating system type, browser version, and error stack information; Step 1.7: Collect the page performance metrics data, JavaScript error messages and access details logs into the local buffer according to the unified session identifier, and wait for the reporting opportunity; Step 1.8: Listen for the beforeunload and visibilitychange events. When the user closes the page or switches to the background, the navigator.sendBeacon interface is called first to send buffered data. If the current environment does not support Beacon, the XMLHttpRequest with keepalive set to true is used as a fallback to complete the transmission.

[0037] The performance monitoring script mentioned above is deployed on the front end of the target application. The performance monitoring script in this invention is client-side execution code used to capture user access behavior, page performance metrics, and network error data in real time. The access details log mentioned above refers to a structured record containing the destination URL, visitor IP address, MAC address, redirection time, and session identifier. The page performance metrics data mentioned above refers to the three core metrics calculated based on the Navigation Timing API: DNS resolution time, SSL connection time, and first byte time (TTFB).

[0038] In some embodiments, the performance monitoring script captures page performance metric data by listening to page lifecycle events, such as... Figure 3 As shown, the specific execution process includes: Step 2.1: Immediately after injection and activation, the script calls the `performance.getEntriesByType` method to obtain the navigation type performance entries for the current page; Step 2.2: Determine if the returned performance entry exists and contains the navigationStart field. If it exists, proceed to step 2.3; otherwise, proceed to step 2.9. Step 2.3: Extract the domainLookupStart and domainLookupEnd timestamps from the performance entries, and calculate the difference between them to obtain the DNS resolution time; Step 2.4: Extract the secureConnectionStart and connectStart timestamps from the performance entries. If secureConnectionStart is valid, calculate the difference between the two to obtain the SSL connection time and proceed to step 2.5. Otherwise, skip the SSL connection time calculation and proceed directly to step 2.5. Step 2.5: Extract the responseStart and navigationStart timestamps from the performance entries, and calculate the difference between them to obtain the first byte time (TTFB). Step 2.6: Determine if the current page has triggered the load event. If it has, proceed to step 2.7; otherwise, register a load event listener and execute step 2.7 after the event is triggered. Step 2.7: Combine the calculated DNS resolution time, SSL connection time, first byte time (TTFB) with the current page URL, user device information, and session identifier into a structured performance metric record; Step 2.8: Record the performance metric in the local performance data buffer and wait for the unified reporting opportunity. Step 2.9, end the page performance metric collection process.

[0039] The page performance metrics data mentioned above refer to the three core metrics—DNS resolution time, SSL connection time, and first byte time (TTFB)—collected and calculated through the browser's Navigation Timing API. The performance metric records mentioned above refer to structured data units containing the above three metrics, page URL, device information, and session identifier. The local performance data buffer mentioned above refers to the memory storage area used to temporarily store performance metric records that have not yet been reported.

[0040] In some embodiments, the performance monitoring script captures JavaScript error information by listening for error events in the page's global object and records the corresponding device information and error stack, such as... Figure 4 As shown, the specific execution process includes: Step 3.1: During the initialization phase, the script calls the window.addEventListener method to register a listener for the unhandledrejection event; Step 3.2: Call the window.onerror property to bind the synchronous JavaScript error handling function; Step 3.3: When an uncaught Promise rejection exception occurs, the unhandledrejection event handler is triggered, and the process proceeds to step 3.4. Step 3.4: Extract the error cause, error message, and associated Promise object from the event object. If the error cause is an Error instance, further extract the stack property as the error stack and proceed to step 3.6. Step 3.5: When a synchronous JavaScript runtime exception occurs, the window.onerror callback is triggered. The error message, error file path, line number, column number, and Error object are obtained from the callback parameters. If the Error object exists, its stack property is extracted as the error stack, and then proceed to step 3.6. Step 3.6: Call the navigator.userAgent interface to parse the current browser version and operating system type; Step 3.7: Combine the error message, error type, error stack, browser version, operating system type, current page URL, and session identifier into a structured JavaScript error message record; Step 3.8: Record the JavaScript error message in the local error data buffer and wait for it to be reported along with the performance metrics data; Step 3.9, end the JavaScript error capture and logging process.

[0041] The JavaScript error information mentioned above refers to the details of an exception triggered and captured by a browser global error event. The JavaScript error information in this invention includes the error type, error message, error stack, and associated client device information. The error stack mentioned above refers to the text information describing the function call chain when the exception occurred. The local error data buffer mentioned above refers to the memory storage area used to temporarily store JavaScript error information records that have not yet been reported.

[0042] In some embodiments, the performance monitoring script records the user's page navigation behavior by hijacking the browser's native navigation mechanism, generating access detail logs containing the destination URL, visitor IP address, and MAC address, such as... Figure 5 As shown, the specific execution process includes: Step 4.1: During the initialization phase, the script saves the original references to the window.history.pushState and replaceState methods; Step 4.2, override the window.history.pushState method to intercept the incoming destination URL parameter before calling the original method; Step 4.3, override the window.history.replaceState method to intercept the incoming destination URL parameter before calling the original method; Step 4.4: When the user triggers a front-end route jump, the newly defined pushState or replaceState method is called, and the process proceeds to step 4.5. Step 4.5: Send a probe request to the server's lightweight interface via navigator.sendBeacon or fetch, and the server will return the public IP address of the current visitor; Step 4.6: Determine if the browser supports WebRTC. If it does, try to obtain the internal network IP address and deduce the MAC address through RTCPeerConnection, and proceed to step 4.7. Otherwise, skip MAC address collection and proceed directly to step 4.7. Step 4.7: Combine the destination URL, visitor IP address, MAC address, redirect timestamp, and current session identifier into a structured access details log; Step 4.8: Write the access details log to the local access log buffer, and wait for it to be reported in batches along with performance metrics data and JavaScript error information; Step 4.9, end the page redirection behavior recording process.

[0043] The aforementioned access detail log refers to a structured record generated by the performance monitoring script when a user navigates to a different page. The access detail log in this invention includes the destination URL, visitor IP address, MAC address, navigation time, and session identifier. The aforementioned local access log buffer refers to a memory storage area used to temporarily store access detail logs that have not yet been reported. The aforementioned session identifier is a unique string generated based on device fingerprint and user behavior, used to identify all operations of the same user within a single access period.

[0044] Furthermore, page performance metrics, JavaScript error messages, and access details logs are reported to the server as raw data.

[0045] In some embodiments, regarding the multi-dimensional aggregation of application access and performance statistics by combining a preset time dimension or a custom time range, the first statistics module is specifically used for: Obtain the time dimension selected by the user. The time dimension can be any one of days, months, or years, or a start and end time range specified by the user. Based on the time dimension, user access behavior data is aggregated by UV and PV to obtain access statistics. In some embodiments, the first statistics module aggregates user access behavior data based on the user-selected time dimension, using unique visitor counts (UV) and page views (PV) to obtain access statistics, such as... Figure 6 As shown, the specific execution process includes: Step 5.1: Filter out all access detail logs that match the specified time dimension from the server's original data storage. Each log contains a session identifier, destination URL, visitor IP address, MAC address, and redirect timestamp. Step 5.2: Determine if the time dimension is a preset granularity. If it is a day, month, or year, proceed to step 5.3. Otherwise, determine if it is a custom time range, process it according to the same logic, and proceed to step 5.3. Step 5.3: Use the session identifier as the primary key to deduplicate the access details logs and count the number of unique sessions, which will be used as the unique visitor count (UV) within that time unit. Step 5.4: Do not deduplicate the access details logs; directly count all log entries within the time unit as the page view count (PV). Step 5.5: Determine whether the current time dimension contains multiple time units. If so, repeat steps 5.3 and 5.4 in chronological order and proceed to step 5.6. Otherwise, proceed directly to step 5.6. Step 5.6: Organize the UV and PV values ​​corresponding to each time unit into structured records using a unified timestamp; Step 5.7: Sort all structured records in ascending order by time to form a complete set of access statistics data; Step 5.8, end the visit count aggregation process.

[0046] The aforementioned visitor statistics refer to a set of structured metrics formed by deduplicating and summing the unique visitor count (UV) and page view count (PV) in the original visit detail logs over a specified time period. The aforementioned session identifier refers to a unique string generated by the performance monitoring script on the front end and reported with the visit detail logs, used to identify the cycle of a single user visit. The aforementioned visit detail logs refer to structured records containing the destination URL, visitor IP address, MAC address, redirect time, and session identifier.

[0047] Based on the time dimension, the page performance metric data is calculated by quantiles according to DNS resolution time, SSL connection time and TTFB to obtain performance metric statistics. In some embodiments, the first statistics module performs quantile calculations on the page performance metric data according to the user-selected time dimension, based on DNS resolution time, SSL connection time, and first byte time (TTFB), to obtain performance metric statistics, such as... Figure 7 As shown, the specific execution process includes: Step 6.1: Extract all page performance metric records that conform to the specified time dimension from the server's original data storage. Each record contains the millisecond values ​​of DNS resolution time, SSL connection time, and first byte time (TTFB). Step 6.2: Determine if the number of extracted performance metric records is greater than zero. If it is greater than zero, proceed to step 6.3; otherwise, skip the subsequent calculations and proceed directly to step 6.8. Step 6.3: Sort the DNS resolution time field in all records in ascending order and calculate the 50th, 75th, and 95th percentile values; Step 6.4: Sort the SSL connection time field in all records in ascending order and calculate the 50th, 75th, and 95th percentile values; Step 6.5: Sort the first byte time TTFB field of all records in ascending order, and calculate the 50th, 75th, and 95th percentile values; Step 6.6: Determine whether the current time dimension contains multiple time units. If so, repeat steps 6.1 to 6.5 for each unit in chronological order and proceed to step 6.7. Otherwise, proceed directly to step 6.7. Step 6.7: Organize the values ​​of DNS resolution time, SSL connection time, and first byte time (TTFB) at each time point for each time unit into structured performance metric statistics using a unified timestamp. Step 6.8, end the performance index quantile calculation process.

[0048] The aforementioned performance metric statistics refer to a set of structured values ​​formed by performing quantile calculations on three types of page performance metrics—DNS resolution time, SSL connection time, and first byte time (TTFB)—at a specified time dimension. The aforementioned page performance metric records refer to structured data units reported by performance monitoring scripts that contain the three core performance metrics. The aforementioned quantile calculation refers to a statistical method of dividing a set of ordered values ​​proportionally and taking values ​​at specific positions. In this invention, quantile calculation is used to measure data distribution characteristics.

[0049] By linking access statistics with performance metric statistics, application access and performance statistics are generated.

[0050] In some embodiments, the first statistics module correlates access statistics with performance metric statistics to form application access and performance statistics, such as... Figure 8 As shown, the specific execution process includes: Step 7.1: Confirm that the visit statistics and performance metrics statistics are organized according to the same time dimension, and that each data item contains a unified timestamp identifier; Step 7.2: Iterate through each record in the visit statistics and extract its timestamp, unique visitors (UV), and page views (PV); Step 7.3: Based on the current timestamp, find the performance quantile record for the corresponding time unit in the performance metric statistics. Step 7.4: Determine if a matching performance quantile record is found. If found, proceed to step 7.5; otherwise, skip the current record and continue processing the next one. Step 7.5: Extract the 95th percentile value of DNS resolution time, the 95th percentile value of SSL connection time, and the 95th percentile value of first byte time (TTFB) from the matched performance quantile records; Step 7.6: Combine the timestamp, UV, PV and the extracted three performance quantile values ​​from the current access record into a single joint structured record; Step 7.7: Determine whether all visit statistics records have been processed. If not, return to step 7.2 to continue the loop; otherwise, proceed to step 7.8. Step 7.8: Sort all the combined structured records in ascending order by timestamp to form a complete set of application access and performance statistics. Step 7.9: End the data association process.

[0051] The application access and performance statistics mentioned above refer to a comprehensive data set formed by structurally associating the UV and PV metrics of user access behavior with the DNS resolution time, SSL connection time, first byte time, and TTFB quantile metrics of page performance at the same time dimension; the joint structured record mentioned above refers to a single data unit containing timestamp, UV, PV, and three performance quantile values.

[0052] In some embodiments, in classifying and statistically analyzing network error data to obtain error analysis statistics, the second statistics module is specifically used for: Filter the HTTP status codes and determine that requests with a status code greater than or equal to 400 are abnormal requests; It should be noted that the HTTP protocol defines a response with a status code less than 400 to indicate that the client request has been successfully received, understood, or processed, including 2xx success classes and 3xx redirection classes. On the other hand, a status code greater than or equal to 400 clearly indicates a client error or server-side abnormality, covering 4xx situations such as invalid client request, authentication failure, and resource not found, as well as 5xx failure types such as internal server error, gateway timeout, and service unavailability. If such requests are included in normal behavior statistics, it will seriously interfere with user experience evaluation and system stability judgment. For example, in an e-commerce application, when a user submits an order, the backend interface returns a 502 status code, indicating that the gateway failed to obtain a valid response from the upstream service. This request should be marked as an abnormal request. Similarly, when the frontend loads product images and receives a 404 status code, it indicates that the resource path is incorrect or the file is missing, which is also an abnormal request. Continuing with the above examples, during the Black Friday promotion, when a large number of users access the checkout page, the backend frequently returns 502 status codes due to excessive load. The second statistics module accurately identifies all 502 responses as abnormal requests through a status code filtering mechanism and incorporates them into the calculation of the total number of errors and the percentage of affected users. Among them, abnormal requests refer to request records in the original network error data with an HTTP status code value greater than or equal to 400, which are used as the core input source for error analysis statistics.

[0053] Count the total number of abnormal requests to obtain the total number of errors; Based on the user identifier associated with the abnormal request, calculate the proportion of users who have made at least one abnormal request out of the total number of active users to obtain the percentage of affected users; The HTTP status codes of abnormal requests are sorted by frequency, and the top five status codes and their frequencies are taken to obtain the TOP5 distribution of abnormal status codes. Extract client environment information from JavaScript error messages and abnormal requests, including operating system type and browser version. Count the number of errors under each environment combination to obtain the distribution of client environments of users who reported errors.

[0054] It should be noted that different terminal devices and browser kernels parse and execute front-end code differently. If only the total number of errors is counted without distinguishing the runtime environment, it will be impossible to identify compatibility issues or rendering anomalies under specific platforms. For example, a piece of syntax may run normally in a new version of Chrome, but trigger a TypeError in an old version of Safari. If error data is not aggregated according to the combination of operating system and browser version, such platform-specific defects will be masked by the overall average, leading to deviations in the direction of repair. For example, in e-commerce applications, some users encounter a "Cannot read property 'length' of undefined" error when clicking the "Add to Cart" button. This error only occurs on iOS devices with the Safari browser and is not reproduced in Android Chrome or Windows Edge.

[0055] Continuing with the example above, during Black Friday sales, the system discovered that a large number of 502 exception requests and JavaScript runtime errors were concentrated in the "iOS + Safari" combination environment. Based on this, the second statistics module concatenated the operating system type and browser version into a unique environment key, and added up the number of JavaScript errors and HTTP exception requests respectively. Finally, it generated a distribution result showing that the "iOS + Safari" error frequency was significantly higher than other combinations, supporting the front-end team to quickly locate compatibility defects and release targeted polyfill patches.

[0056] The aforementioned error reporting user client environment distribution refers to the error frequency statistics formed by aggregating all JavaScript errors and HTTP exception requests by operating system type and browser version within a specified time dimension. The error reporting user client environment distribution in this invention is used to reveal the distribution characteristics of front-end exceptions in different terminal environments.

[0057] In some embodiments, in determining the target optimization strategy based on application access and performance statistics and error analysis statistics, the third determining module is specifically used for: Obtain the preset performance baseline threshold and error rate alarm threshold; The third determination module obtains preset performance baseline thresholds and error rate alarm thresholds, such as Figure 9 As shown, the specific execution process includes: Step 8.1: Read the performance baseline threshold configuration item from the system configuration center. This configuration item contains the 95th percentile upper limit values ​​for three metrics: DNS resolution time, SSL connection time, and first byte time (TTFB). Step 8.2: Determine whether the performance baseline threshold configuration item completely includes the threshold definitions of the three metrics. If it is complete, proceed to step 8.3; otherwise, terminate the process and record the configuration missing exception. Step 8.3: Read the error rate alarm threshold configuration item from the system configuration center. This configuration item defines the upper limit percentage of alarm triggering based on the percentage of affected users. Step 8.4: Determine whether the error rate alarm threshold is a valid positive number and does not exceed 100%. If it is valid, proceed to step 8.5; otherwise, terminate the process and record the threshold as invalid or abnormal. Step 8.5: Load the performance baseline threshold and error rate alarm threshold into the policy judgment context memory area; Step 8.6: Verify that all threshold units and data types conform to the expected format. If they do, proceed to step 8.7; otherwise, return to step 8.1 to reload the configuration. Step 8.7: Complete threshold initialization for subsequent comparison with actual statistical data; Step 8.8: End the threshold acquisition process.

[0058] The aforementioned performance baseline threshold refers to the pre-set acceptable upper limit for three types of page performance indicators: DNS resolution time, SSL connection time, and first byte time (TTFB). The performance baseline threshold in this invention is usually determined based on historical stable period data or business SLA requirements. The aforementioned error rate alarm threshold refers to the preset alarm triggering upper limit of the percentage of affected users. The error rate alarm threshold in this invention is used to measure whether the impact of an anomaly on the user group reaches a level that requires intervention. The aforementioned policy judgment context refers to the runtime environment that temporarily stores thresholds and statistical data in memory to support condition comparison.

[0059] The performance metric statistics are compared with the performance baseline threshold. If at least one performance metric exceeds the performance baseline threshold, a performance optimization strategy is triggered. In some embodiments, the third determining module compares the performance metric statistics with the performance baseline threshold. If at least one performance metric exceeds the performance baseline threshold, a performance optimization strategy is triggered, such as... Figure 10 As shown, the specific execution process includes: Step 9.1: Extract the 95th percentile value of DNS resolution time, the 95th percentile value of SSL connection time, and the 95th percentile value of first byte time (TTFB) from the performance metric statistics. Step 9.2: Compare the 95th percentile value of DNS resolution time with the preset DNS resolution time performance baseline threshold. If it exceeds the threshold, proceed to step 9.5; otherwise, proceed to step 9.3. Step 9.3: Compare the 95th percentile value of the SSL connection time with the preset SSL connection time performance baseline threshold. If the value exceeds the threshold, proceed to step 9.5; otherwise, proceed to step 9.4. Step 9.4: Compare the 95th percentile value of the first byte time TTFB with the preset TTFB performance baseline threshold. If it exceeds the threshold, proceed to step 9.5; otherwise, proceed to step 9.8. Step 9.5: Determine if performance degradation exists, and initiate the performance optimization strategy generation process; Step 9.6: Match the corresponding optimization measure template according to the type of the over-limit indicator. If the DNS resolution time exceeds the limit, generate the suggestion of "optimize DNS service provider configuration or enable local caching"; if the SSL connection time exceeds the limit, generate the suggestion of "upgrade TLS protocol version or reuse connection session"; if the first byte time TTFB exceeds the limit, generate the suggestion of "enable CDN edge caching, aggregate backend interface or compress first screen resources". Step 9.7: Combine one or more matching optimization suggestions into a structured performance optimization strategy, mark the strategy type as "performance optimization strategy", and associate it with the current time dimension and application identifier; Step 9.8: End the performance comparison and policy triggering process.

[0060] The aforementioned performance optimization strategy refers to a set of technical improvement schemes automatically generated by the system to improve the loading efficiency and response speed of the target application when any one of the page performance metric statistics exceeds the preset performance baseline threshold. The aforementioned performance metric statistics refer to a set of structured values ​​including DNS resolution time, SSL connection time, first byte time (TTFB) at a specified percentile. The aforementioned performance baseline threshold refers to the pre-set acceptable upper limit value for the above three indicators. The performance baseline threshold in this invention is used to measure whether the front-end performance meets the user experience requirements.

[0061] The percentage of affected users is compared with the error rate alarm threshold. If the percentage of affected users exceeds the error rate alarm threshold, a troubleshooting strategy is triggered. In some embodiments, the third determining module compares the percentage of affected users with the error rate alarm threshold. If the percentage of affected users exceeds the error rate alarm threshold, a fault investigation strategy is triggered, such as... Figure 11 As shown, the specific execution process includes: Step 10.1: Extract the percentage of affected users in the current time dimension from the error analysis statistics; Step 10.2: Read the percentage upper limit value in the preset error rate alarm threshold configuration item; Step 10.3: Determine whether the percentage of affected users is greater than the error rate alarm threshold. If yes, proceed to step 10.4; otherwise, proceed to step 10.9. Step 10.4: Determine that the scope of the abnormal impact has reached the level requiring intervention, and initiate the fault investigation strategy generation process; Step 10.5: Call the TOP5 distribution data of abnormal status codes output by the second statistics module to identify the most frequently occurring HTTP status codes and their associated interface paths; Step 10.6: Call the error reporting user client environment distribution data output by the second statistics module to identify the combination of operating system type and browser version where the error occurs most frequently; Step 10.7: Based on the combination of high-frequency status codes and high-incidence environments, generate structured fault investigation tasks, including specific instructions such as "focus on investigating 502 errors returned by the payment gateway", "verify script compatibility in the iOS Safari environment", and "check for 404 missing paths on the static resource server". Step 10.8: Encapsulate all troubleshooting tasks into a troubleshooting strategy, mark the strategy type as "troubleshooting strategy", and bind the current application identifier and time dimension; Step 10.9: End the error impact comparison and policy triggering process.

[0062] The aforementioned troubleshooting strategy refers to a set of operational guidelines automatically generated by the system when the proportion of affected users exceeds the preset error rate alarm threshold. This strategy is used to locate and repair the root cause of network or script errors encountered by a large number of users. The aforementioned proportion of affected users refers to the ratio of the number of unique users who have made at least one HTTP abnormal request within a specified time period to the total number of active users on that day.

[0063] If both performance optimization and troubleshooting strategies are triggered simultaneously, a comprehensive optimization suggestion will be generated as the target optimization strategy.

[0064] In some embodiments, when the third determining module simultaneously triggers the performance optimization strategy and the fault diagnosis strategy, it generates a comprehensive optimization suggestion as the target optimization strategy, such as... Figure 12 As shown, the specific execution process includes: Step 11.1: Check if a performance optimization strategy has been generated in the current cycle. If yes, proceed to step 11.2; otherwise, end the process. Step 11.2: Check if a troubleshooting strategy has been generated in the current cycle. If yes, proceed to step 11.3; otherwise, end the process. Step 11.3: Retrieve the types of out-of-limit performance indicators and corresponding optimization measures included in the performance optimization strategy; Step 11.4: Retrieve the high-frequency abnormal status codes, high-incidence client environment combinations, and associated interface paths contained in the fault diagnosis strategy; Step 11.5: Analyze whether there is a spatiotemporal overlap between performance degradation and error distribution. If both are concentrated in the same interface or user path, proceed to step 11.6; otherwise, merge according to the main path with the strongest correlation. Step 11.6: Generate joint action items based on cross-evidence, including "Implement backend cache degradation for payment interface to reduce TTFB", "Inject syntax compatibility polyfill for Safari environment to fix JavaScript errors", "Increase CDN edge node origin timeout threshold to alleviate 502 errors", and "Temporarily limit non-core requests to ensure main link stability". Step 11.7: Integrate all joint action items into a structured comprehensive optimization suggestion, mark the strategy type as "comprehensive optimization suggestion", and bind the application identifier, time dimension and priority label; Step 11.8: End the comprehensive strategy generation process.

[0065] The aforementioned comprehensive optimization suggestion refers to a combined action guide automatically generated by the system that integrates performance tuning measures and root cause repair solutions when performance optimization strategies and fault diagnosis strategies are triggered simultaneously. The aforementioned target optimization strategy refers to a structured strategy entity that is ultimately output to guide the operation and maintenance and development teams in performing improvement operations. The aforementioned spatiotemporal intersection refers to the overlap between performance degradation and high error incidence in the same time window and on the same user path.

[0066] In some embodiments, the system further includes an alarm linkage module, which is used for: Once the third determination module determines the target optimization strategy, it generates corresponding alarm information based on the type of the target optimization strategy. Alarm information is pushed to the preset operation and maintenance terminal, and multi-dimensional visualization charts and anomaly distribution visualization charts are exported as attachments along with the alarm information.

[0067] In some embodiments, the first visualization module and the second visualization module are specifically used to transform the statistical analysis results into intuitive, interactive graphical views.

[0068] It is important to distinguish that the first visualization module focuses on presenting the time-series trends of traffic and performance metrics, while the second visualization module focuses on the structured display of error distribution and user impact dimensions.

[0069] Specifically, the first visualization module is a graphical generation unit used to transform access statistics and performance indicator statistics into time-series trend charts. The output of the first visualization module includes UV / PV trends, performance quantile curves, and performance baseline reference lines. The second visualization module is a graphical generation unit used to transform error analysis statistics into error distribution and user impact views. The output of the second visualization module includes an abnormal status code percentage chart, a client environment error heatmap, and a time-series chart of the percentage of affected users.

[0070] Example 2, refer to Figure 13 This embodiment also provides a browser-based multidimensional application access and performance data collection and analysis method, including: Acquire user access behavior data, page performance index data, and network error data of the target application dynamically collected via browser-side scripts; User access behavior data includes unique visitors (UV) and page views (PV); Page performance metrics include DNS resolution time, SSL connection time, and first byte time (TTFB). Network error data includes HTTP status codes and JavaScript error messages; Based on user access behavior data and page performance metrics data, multi-dimensional aggregation is performed using preset time dimensions or custom time ranges to obtain application access and performance statistics. Multidimensional visualization charts are generated and displayed based on application access and performance statistics. These charts include line charts to show access trends and bar charts to show the distribution of performance bottlenecks. The network error data is classified and statistically analyzed to obtain error analysis statistics. The error analysis statistics include the total number of errors, the percentage of affected users, the distribution of the top 5 abnormal status codes, and the distribution of the client environment of users who reported errors. Anomaly distribution visualization charts are generated and displayed based on error analysis statistics. These charts include bar charts to show the distribution of error status codes and pie charts to show the distribution of error user client environments. Based on application access and performance statistics and error analysis statistics, target optimization strategies are determined to improve the stability of the target application and user experience.

[0071] The above-mentioned unit modules can be embedded in the processor of the electronic device in hardware form or independent of it, or they can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0072] This embodiment also provides an electronic device, which can be a terminal, and its internal structure diagram can be as follows: Figure 13As shown, the electronic device includes a processor, memory, communication interface, display screen, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a browser-based multi-dimensional application access and performance data acquisition and analysis method. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the device's casing, or an external keyboard, touchpad, or mouse.

[0073] This embodiment also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, it performs the following steps: Acquire user access behavior data, page performance index data, and network error data of the target application dynamically collected via browser-side scripts; User access behavior data includes unique visitors (UV) and page views (PV); Page performance metrics include DNS resolution time, SSL connection time, and first byte time (TTFB). Network error data includes HTTP status codes and JavaScript error messages; Based on user access behavior data and page performance metrics data, multi-dimensional aggregation is performed using preset time dimensions or custom time ranges to obtain application access and performance statistics. Multidimensional visualization charts are generated and displayed based on application access and performance statistics. These charts include line charts to show access trends and bar charts to show the distribution of performance bottlenecks. The network error data is classified and statistically analyzed to obtain error analysis statistics. The error analysis statistics include the total number of errors, the percentage of affected users, the distribution of the top 5 abnormal status codes, and the distribution of the client environment of users who reported errors. Anomaly distribution visualization charts are generated and displayed based on error analysis statistics. These charts include bar charts to show the distribution of error status codes and pie charts to show the distribution of error user client environments. Based on application access and performance statistics and error analysis statistics, target optimization strategies are determined to improve the stability of the target application and user experience.

[0074] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0075] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0076] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A browser-based multidimensional application access and performance data acquisition and analysis system, characterized in that, The system includes a data acquisition module, a first statistics module, a first visualization module, a second statistics module, a second visualization module, and a third determination module, wherein: The data acquisition module is used to dynamically collect user access behavior data, page performance index data, and network error data of the target application through browser-side scripts; The user access behavior data includes unique visitors (UV) and page views (PV); The page performance metrics data include DNS resolution time, SSL connection time, and first byte time (TTFB). The network error data includes HTTP status codes and JavaScript error messages; The first statistics module is used to perform multi-dimensional aggregation based on the user access behavior data and the page performance index data, combined with a preset time dimension or a custom time range, to obtain application access and performance statistics data. The first visualization module is used to generate and display multi-dimensional visualization charts based on the application access and performance statistics. The multi-dimensional visualization charts include line charts for presenting access trends and bar charts for presenting the distribution of performance bottlenecks.

2. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 1, characterized in that, Also includes: The second statistics module is used to classify and statistically analyze the network error data to obtain error analysis statistics. The error analysis statistics include the total number of errors, the percentage of affected users, the top 5 distribution of abnormal status codes, and the distribution of client environments of users who reported errors. The second visualization module is used to generate and display anomaly distribution visualization charts based on the error analysis statistics. The anomaly distribution visualization charts include a bar chart for presenting the distribution of abnormal status codes and a pie chart for presenting the distribution of the user client environment of the error. The third determining module is used to determine the target optimization strategy based on the application access and performance statistics and the error analysis statistics, so as to improve the stability and user experience of the target application.

3. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 2, characterized in that, Regarding the dynamic collection of user access behavior data, page performance index data, and network error data of the target application via browser-side scripts, the data collection module is specifically used for: Inject a performance monitoring script when the target application's page loads; The performance monitoring script listens for page lifecycle events and captures page performance metrics data. The performance monitoring script listens for error events in the page's global object, captures JavaScript error information, and records the corresponding device information and error stack. The performance monitoring script records the user's page navigation behavior and generates access detail logs containing the destination URL, visitor IP address, and MAC address. The page performance metrics data, the JavaScript error information, and the access details log are reported to the server as raw data.

4. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 3, characterized in that, Regarding the aggregation of application access and performance statistics by combining preset time dimensions or custom time ranges into multiple dimensions, the first statistics module is specifically used for: Obtain the time dimension selected by the user, which includes any one of days, months, and years, or a start and end time range specified by the user; Based on the time dimension, the user access behavior data is aggregated by UV and PV to obtain access statistics. Based on the time dimension, the page performance index data is calculated by quantiles according to DNS resolution time, SSL connection time and TTFB to obtain performance index statistics. The access statistics are correlated with the performance metrics statistics to form the application access and performance statistics.

5. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 4, characterized in that, In the process of classifying and statistically analyzing the network error data to obtain error analysis statistics, the second statistical module is specifically used for: The HTTP status codes are filtered to determine that requests with a status code greater than or equal to 400 are abnormal requests. The total number of errors is obtained by counting the total number of abnormal requests; Based on the user identifier associated with the abnormal request, the proportion of users who have made at least one abnormal request to the total number of active users is calculated to obtain the proportion of affected users. The HTTP status codes of the abnormal requests are sorted by frequency, and the top five status codes and their frequencies are taken to obtain the TOP5 distribution of the abnormal status codes. Extract the JavaScript error information and the client environment information contained in the abnormal request. The client environment information includes the operating system type and browser version. Count the number of errors under each environment combination to obtain the client environment distribution of the user who reported the error.

6. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 5, characterized in that, In determining the target optimization strategy based on the application access and performance statistics and the error analysis statistics, the third determining module is specifically used for: Obtain the preset performance baseline threshold and error rate alarm threshold; The performance metric statistics are compared with the performance baseline threshold. If at least one performance metric exceeds the performance baseline threshold, a performance optimization strategy is triggered. The percentage of affected users is compared with the error rate alarm threshold. If the percentage of affected users exceeds the error rate alarm threshold, a troubleshooting strategy is triggered. If both the performance optimization strategy and the fault diagnosis strategy are triggered simultaneously, a comprehensive optimization suggestion is generated as the target optimization strategy.

7. The browser-based multidimensional application access and performance data acquisition and analysis system as described in claim 6, characterized in that, The system also includes an alarm linkage module, which is used for: Once the third determining module determines the target optimization strategy, it generates corresponding alarm information based on the type of the target optimization strategy. The alarm information is pushed to a preset operation and maintenance terminal, and the multi-dimensional visualization chart and the abnormal distribution visualization chart can be exported as attachments along with the alarm information.

8. A browser-based method for collecting and analyzing multi-dimensional application access and performance data, using the system described in any one of claims 1 to 7, characterized in that, include: Acquire user access behavior data, page performance index data, and network error data of the target application dynamically collected via browser-side scripts; The user access behavior data includes unique visitors (UV) and page views (PV); The page performance metrics data include DNS resolution time, SSL connection time, and first byte time (TTFB). The network error data includes HTTP status codes and JavaScript error messages; Based on the user access behavior data and the page performance index data, multi-dimensional aggregation is performed in combination with a preset time dimension or a custom time range to obtain application access and performance statistics. Based on the application access and performance statistics, multi-dimensional visualization charts are generated and displayed, including line charts to present access trends and bar charts to present the distribution of performance bottlenecks. The network error data is classified and statistically analyzed to obtain error analysis statistics; the error analysis statistics include the total number of errors, the percentage of affected users, the distribution of the top 5 abnormal status codes, and the distribution of the client environment of the users who reported errors. An anomaly distribution visualization chart is generated and displayed based on the error analysis statistics. The anomaly distribution visualization chart includes a bar chart to present the distribution of anomaly status codes and a pie chart to present the distribution of the user client environment of the user who reported the error. Based on the application access and performance statistics and the error analysis statistics, a target optimization strategy is determined to improve the stability and user experience of the target application.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the browser-based multidimensional application access and performance data acquisition and analysis method as described in claim 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the browser-based multidimensional application access and performance data collection and analysis method described in claim 8.

Images