Real-time pass rate matrix monitoring system and method

Abstract

The application discloses a real-time pass rate matrix monitoring system and method, relates to the technical field of test quality monitoring, and comprises the following steps: receiving job records of a task execution resource pool, establishing performance statistical data according to resource identifiers, identifying and correcting data consistency conflicts among data sources through multi-source data fusion, updating total task quantity and completion quantity, and outputting real-time quality indicators and updated statistical information; constructing a management multi-dimensional matrix of resource dimensions, index dimensions and state dimensions according to risk assessment marks and fault cause information, mapping a risk-free state into a safety level identifier, and generating a differentiated risk level identifier based on fault cause information mapping. The application constructs a management multi-dimensional matrix, generates a differentiated risk level identifier based on fault cause information mapping, triggers resource scheduling based on the differentiated risk level identifier, improves the pertinence of scheduling strategies and the accuracy of abnormal processing, and guarantees stable operation of the task execution resource pool.

Description

Technical Field

[0001] This invention relates to the field of test quality monitoring technology, and in particular to a real-time pass rate matrix monitoring system and method. Background Technology

[0002] As distributed test clusters grow in scale, real-time pass rate monitoring has become a crucial technology for ensuring task execution quality. Existing methods collect statistics from multiple job records, use comparisons to achieve status monitoring and visualization, and are widely used in the field of automated testing.

[0003] However, existing methods lack consistency conflict correction when fusing multi-source data, leading to statistical benchmark distortion. Furthermore, risk assessment is typically based on a single node, failing to generate differentiated risk levels based on fault attribution. This results in coarse-grained resource scheduling strategies, hindering refined management and impacting overall operational stability and resource utilization. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a real-time pass rate matrix monitoring method to solve the problem of inadequate control caused by missing consistency conflict correction for multi-source data and insufficient risk level differentiation mapping.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a real-time pass rate matrix monitoring method, which includes: receiving job records from a task execution resource pool, establishing performance statistics data according to resource identifiers, identifying and correcting data consistency conflicts between data sources through multi-source data fusion, updating the total task volume and completion volume, and outputting real-time quality indicators and updated statistical information; extracting historical quality indicator sequences based on the updated statistical information, calculating statistical feature values ​​for the historical quality indicator sequences and generating performance early warning benchmarks, comparing real-time quality indicators with performance early warning benchmarks to obtain risk assessment indicators; when the risk assessment indicator indicates an early warning, extracting abnormal execution features and classifying and encoding them to generate fault attribution information; constructing a multi-dimensional management matrix with resource, indicator, and status dimensions based on the risk assessment indicator and fault attribution information, mapping risk-free status to safety level identifiers, and generating differentiated risk level identifiers based on the fault attribution information; refreshing the real-time management status interface according to the multi-dimensional management matrix, matching differentiated risk level identifiers and safety level identifiers, and triggering resource scheduling, task reallocation, and anomaly handling processes.

[0008] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the task execution resource pool refers to a resource cluster composed of multiple nodes with independent network location information, and each node corresponds to a unique resource identifier for performance tracking.

[0009] The resource identifier includes network location information, and the job record includes the total number of business tasks generated by the node in parallel processing within the management and statistical period and the number of tasks completed.

[0010] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the specific steps for outputting real-time quality indicators and updated statistical information are as follows:

[0011] Based on the job records of the task execution resource pool, data from each node is extracted by resource identifier and classified and organized for performance management to obtain performance statistics data distinguished by resource identifier;

[0012] Extract multi-source business information from performance statistics, identify and merge data consistency conflicts, and obtain updated total task volume and completion statistics.

[0013] Using the updated total task volume and completion statistics, calculate the real-time quality indicators corresponding to each resource identifier to obtain the updated statistics.

[0014] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the historical quality indicator sequence refers to a performance trend analysis set composed of real-time quality indicators corresponding to each resource identifier in multiple consecutive statistical periods in the past, arranged in chronological order according to the management time period.

[0015] The real-time quality indicator is the ratio of the number of tasks completed corresponding to each resource identifier to the total number of tasks, i.e., the pass rate.

[0016] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the specific steps for obtaining the risk assessment indicator are as follows:

[0017] Based on the updated statistical information, the management time window extraction method is used to collect historical quality indicator data and obtain a historical quality indicator sequence arranged in chronological order.

[0018] Statistical characteristics are calculated on historical quality indicator sequences to obtain the mean and fluctuation range. The warning interval is determined based on the mean and fluctuation range and used as the performance warning benchmark.

[0019] Based on the performance early warning benchmark, risk assessment indicators are obtained by comparing and analyzing the differences between real-time quality indicators and the performance early warning benchmark.

[0020] As a preferred embodiment of the real-time throughput matrix monitoring method of the present invention, the specific steps for generating fault attribution information are as follows:

[0021] Based on risk assessment indicators, abnormal execution features are extracted. By collecting pass rate data, task execution times, and completion quantities during the work process and reorganizing them, an abnormal feature dataset is obtained.

[0022] The abnormal execution features in the abnormal feature dataset are classified and encoded to obtain an abnormal classification and encoding sequence.

[0023] By comparing and classifying the anomaly classification coding sequence with the anomaly classification coding of each node in the task execution resource pool, fault attribution information is obtained. Nodes with matching codes are identified as fault sources of the same type and the attribution information is output.

[0024] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the management multidimensional matrix refers to a resource management data structure with the resource identifier as the index dimension and the real-time quality indicator value and status identifier as the data dimension. The status identifier is mapped to a security level identifier and a differentiated risk level identifier based on the comparison between the real-time quality indicator value and the performance warning benchmark, thus forming a real-time management matrix of resources and pass rate.

[0025] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the specific steps for generating differentiated risk level identifiers are as follows:

[0026] Based on real-time quality indicators, the total task volume and task completion volume corresponding to resource identifiers are reorganized to obtain a multi-dimensional data set corresponding to resources and pass rates.

[0027] Based on the multidimensional dataset, resource identifiers and real-time quality indicators are correlated to obtain a multidimensional management matrix.

[0028] By managing a multi-dimensional matrix for state mapping, risk-free states are mapped to safety level identifiers, and risky states are mapped to generate differentiated risk level identifiers based on fault attribution information.

[0029] As a preferred embodiment of the real-time pass rate matrix monitoring method of the present invention, the specific steps of triggering resource scheduling, task reallocation, and exception handling are as follows:

[0030] Based on the management multidimensional matrix, data mapping processing is performed, and color gradient mapping is applied to status identifiers and differentiated risk level identifiers and bound to interface elements to obtain a refreshed real-time management status interface.

[0031] Based on the refreshed real-time management situation interface, combined with differentiated risk level identifiers and security level identifiers, strategy matching is performed to generate resource scheduling and task reallocation control instructions.

[0032] The exception handling process is executed according to the task reassignment control instructions, which completes resource scheduling, task reassignment and triggering of exception handling process.

[0033] Secondly, this invention provides a real-time pass rate matrix monitoring system, comprising: a statistics maintenance module, used to receive job records from the task execution resource pool, establish performance statistics data according to resource identifiers, identify and correct data consistency conflicts between data sources through multi-source data fusion, update the total task volume and completion volume, and output real-time quality indicators and updated statistical information; an early warning analysis module, used to extract historical quality indicator sequences based on updated statistical information, calculate statistical feature values ​​for historical quality indicator sequences and generate performance early warning benchmarks, compare real-time quality indicators with performance early warning benchmarks, and obtain risk assessment indicators; a fault clustering module, used to extract abnormal execution features and classify and encode them when the risk assessment indicator indicates an early warning, and generate fault attribution information; a matrix generation module, used to construct a multi-dimensional management matrix with resource dimension, indicator dimension, and status dimension based on risk assessment indicators and fault attribution information, map risk-free status to safety level identifiers, and generate differentiated risk level identifiers based on fault attribution information; and a visualization and control module, used to refresh the real-time management status interface according to the multi-dimensional management matrix, match differentiated risk level identifiers and safety level identifiers, and trigger resource scheduling, task reallocation, and anomaly handling processes.

[0034] The beneficial effects of this invention are as follows: by constructing a multi-dimensional management matrix, generating differentiated risk level identifiers based on fault attribution information, and triggering resource scheduling based on the differentiated risk level identifiers, the targeting of scheduling strategies and the accuracy of anomaly handling are improved, ensuring the stable operation of the task execution resource pool. Attached Figure Description

[0035] 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.

[0036] Figure 1 This is a flowchart of a real-time pass rate matrix monitoring method.

[0037] Figure 2 This is a schematic diagram of a real-time pass rate matrix monitoring system.

[0038] Figure 3 This is a flowchart for real-time pass rate matrix monitoring.

[0039] Figure 4 A flowchart for visualizing the real-time pass rate matrix. Detailed Implementation

[0040] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0041] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0043] Reference Figures 1-4 This is one embodiment of the present invention, which provides a real-time throughput matrix monitoring method, including the following steps:

[0044] S1: Receives job records from the task execution resource pool, establishes performance statistics according to resource identifiers, identifies and corrects data consistency conflicts between data sources through multi-source data fusion, updates the total task volume and completion volume, and outputs real-time quality indicators and updated statistical information.

[0045] S1.1: The task execution resource pool refers to a resource cluster composed of multiple nodes with independent network location information. Each node corresponds to a unique resource identifier for performance tracking.

[0046] Resource identifiers include network location information, and job records include the total number of business tasks generated by the node in parallel processing within the management and statistical period and the number of tasks completed.

[0047] Based on the job records of the task execution resource pool, data from each node is extracted by resource identifier and classified and organized for performance management to obtain performance statistics data distinguished by resource identifier.

[0048] Furthermore, the system receives job records from the task execution resource pool. These records include the total number of business tasks generated by the nodes in parallel processing within the management and statistical period, as well as the number of tasks completed. Data from each node is extracted using resource identifiers, which include network location information. Statistical locks are used to protect node data updates. Total task volume and completion statistics are maintained independently according to resource identifiers. Performance management is then performed, and performance statistics are obtained by differentiating them by resource identifiers.

[0049] Specifically, statistical locking refers to a method of mutual exclusion control for statistical processing corresponding to the same resource identifier when multiple nodes report job records in parallel and update the total task volume and completion statistics at the same time. This ensures that only one update action is allowed to write the corresponding statistical information at any given time, avoiding data overwriting, duplicate accumulation and statistical errors caused by concurrent updates, and ensuring the accuracy of performance statistics maintained by resource identifier.

[0050] Performance management classification and organization refers to the process of collecting data from each node extracted through resource identifiers according to the resource identifiers, classifying and organizing the total number of business tasks and the number of completed tasks corresponding to each resource identifier, and forming performance statistics data distinguished by resource identifiers.

[0051] S1.2: Extract multi-source business information from performance statistics, identify and merge data consistency conflicts, and obtain updated total task volume and completion statistics.

[0052] Furthermore, based on the performance statistics data differentiated by resource identifier, the total number of business tasks and the number of completed tasks corresponding to each resource identifier are extracted from different data sources within the same management statistical period to obtain multi-source business information (total task volume and completion volume of business tasks at each node). Data consistency conflict identification is performed on the multi-source business information, and data consistency conflicts between data sources are corrected by conflict fingerprint extraction. The corrected total number of business tasks and the number of completed tasks are merged according to resource identifier, and the merged total number of business tasks and the number of completed tasks are updated to obtain updated total task volume and completion volume statistics.

[0053] Specifically, consistency conflict identification refers to the process of comparing and analyzing the corresponding content of total task volume and completion volume in multi-source business information to identify inconsistent records between different data sources.

[0054] Fusion processing refers to the process of correcting and integrating multi-source business information after completing consistency conflict identification, in order to obtain updated total task volume and completion statistics.

[0055] S1.3: Calculate the real-time quality indicators corresponding to each resource identifier using the updated total task volume and completion statistics to obtain the updated statistics.

[0056] Furthermore, based on the total task volume and completion volume statistics obtained after data consistency conflict identification and fusion processing, the data is collected and organized according to resource identifiers. The pass rate is determined by the ratio of the task completion volume corresponding to each resource identifier to the total task volume. The pass rate is used as the real-time quality indicator for the corresponding resource identifier. The resource identifier, total task volume, completion volume and real-time quality indicator are correlated to obtain updated statistical information.

[0057] The formula for real-time quality indicator fusion calculation is as follows:

[0058] ;

[0059] in, Representing the Real-time quality metrics for each resource identifier Resource Identifier Index Representing the The total number of business tasks for each data source Represents the data source sequence number. Represents the number of data sources. Representing the The resource identifier is in the first The pass rate of each data source.

[0060] Association processing refers to the process of matching and aggregating resource identifiers with corresponding real-time quality indicator values, total task volume, task completion volume, status identifiers, security level identifiers, and differentiated risk level identifiers to form a multi-dimensional management matrix and a real-time management matrix of resources and pass rates.

[0061] S2: Extract historical quality indicator sequences based on updated statistical information, calculate statistical feature values ​​for historical quality indicator sequences and generate performance early warning benchmarks, compare real-time quality indicators with performance early warning benchmarks, and obtain risk assessment indicators.

[0062] S2.1: The historical quality indicator sequence refers to a performance trend analysis set composed of real-time quality indicators corresponding to each resource identifier within multiple consecutive statistical periods in the past, arranged in chronological order according to the management time period.

[0063] The real-time quality metric is the ratio of the number of tasks completed for each resource identifier to the total number of tasks, i.e., the pass rate.

[0064] Based on the updated statistical information, the management time window extraction method is used to collect historical quality indicator data and obtain a historical quality indicator sequence arranged in chronological order.

[0065] Furthermore, based on the updated statistical information, the management time window extraction method is adopted to collect historical quality indicator data for the real-time quality indicators corresponding to each resource identifier in multiple consecutive statistical periods in the past. The real-time quality indicators collected in multiple consecutive statistical periods in the past are arranged in chronological order according to the management time period to obtain a historical quality indicator sequence arranged in chronological order.

[0066] It should be noted that the management time window extraction method refers to a processing method that, based on updated statistical information, determines the collection range of historical quality indicator data according to the management time cycle, and within the time range, continuously extracts, arranges and aggregates the real-time quality indicators corresponding to each resource identifier in chronological order to form a historical quality indicator sequence.

[0067] S2.2: Perform statistical characteristic calculations on the historical quality indicator series to obtain the mean and fluctuation range, determine the warning interval based on the mean and fluctuation range, and use the warning interval as the performance warning benchmark.

[0068] Furthermore, when performing statistical feature calculations on the historical quality indicator series, the real-time quality indicators in the historical quality indicator series are processed according to the chronological order of the management time cycle. The real-time quality indicators corresponding to the historical quality indicator series are aggregated and calculated to obtain the mean. Then, based on the changes of the real-time quality indicators in the historical quality indicator series around the mean, statistical feature calculations are performed to obtain the fluctuation range. Based on the mean and the fluctuation range, the interval is determined to obtain the warning interval, which is used as the performance warning benchmark.

[0069] Specifically, the warning interval refers to the normal fluctuation range of real-time quality indicators determined by statistically calculating the historical quality indicator sequence and combining the mean and fluctuation range. For example, based on the management time window extraction method, real-time quality indicators corresponding to each resource identifier in multiple consecutive statistical periods are collected. After forming a historical quality indicator sequence according to the order of management time periods, the historical quality indicator sequence is summarized and calculated to obtain the mean. The fluctuation range is obtained based on the changes of each real-time quality indicator in the historical quality indicator sequence around the mean.

[0070] Determining the range based on the mean and fluctuation range refers to summarizing the real-time quality indicators in the historical quality indicator sequence after extracting and forming a historical quality indicator sequence according to the management time period, obtaining the mean, and statistically analyzing the degree of deviation of each real-time quality indicator from the mean to obtain the fluctuation range.

[0071] The fluctuation range is derived from the statistical information on the deviation of each real-time quality indicator from the mean in the historical quality indicator sequence. Based on the management time window extraction method, the real-time quality indicators corresponding to each resource identifier are collected in the past multiple consecutive statistical periods. After forming a historical quality indicator sequence according to the chronological order of the management time period, the historical quality indicator sequence is summarized to obtain the mean. The fluctuation range is determined according to the deviation of each real-time quality indicator from the mean. The mean minus the fluctuation range is used as the lower limit of the warning interval, and the mean plus the fluctuation range is used as the upper limit of the warning interval.

[0072] The warning interval is determined based on the mean and fluctuation range, and is used as the performance warning benchmark. When the real-time quality indicator corresponding to the current resource identifier is within the warning interval, it indicates that the pass rate of the corresponding node in the task execution resource pool is in a normal fluctuation state. If the real-time quality indicator corresponding to the current resource identifier is lower than the lower limit of the warning interval, it indicates that the corresponding node has an abnormal pass rate risk.

[0073] S2.3: Based on the performance early warning benchmark, risk assessment indicators are obtained by comparing and analyzing the differences between real-time quality indicators and the performance early warning benchmark.

[0074] Furthermore, based on the performance early warning benchmark, real-time quality indicators corresponding to each resource identifier are extracted from the updated statistical information. The real-time quality indicators corresponding to each resource identifier are compared and analyzed with the performance early warning benchmark, and the difference assessment information is evaluated. When the real-time quality indicator is lower than the performance early warning benchmark, a risk assessment flag for the corresponding resource identifier is obtained to indicate an early warning. When the real-time quality indicator is higher than the performance early warning benchmark, a risk assessment flag for the corresponding resource identifier is obtained without indicating an early warning. The difference assessment information is then classified and organized according to the resource identifier for performance management to obtain the risk assessment flag.

[0075] S3: When the risk assessment indicator indicates an early warning, extract abnormal execution characteristics, classify and encode them, and generate fault attribution information.

[0076] S3.1: Based on risk assessment flags, abnormal execution feature extraction is triggered by collecting pass rate data, task execution times and completion quantity information during the work process and reorganizing them to obtain an abnormal feature dataset.

[0077] Furthermore, when the risk assessment indicator indicates an early warning, based on the job records and updated statistical information of the task execution resource pool, abnormal execution features are extracted from the nodes corresponding to each resource identifier. Pass rate data, task execution counts, and completion quantity information are collected during the work process. The pass rate data, task execution counts, and completion quantity information are then classified, organized, and reorganized according to the resource identifier for performance management. Each resource identifier is then assigned to the pass rate data, task execution counts, and completion quantity information to obtain an abnormal feature dataset.

[0078] Specifically, the extraction of abnormal execution features involves comparing and analyzing real-time quality indicators with the performance early warning benchmark based on the performance early warning benchmark and obtaining risk assessment indicators. When the risk assessment indicators indicate an early warning, the pass rate data, task execution count, and completion quantity information of the corresponding resource identifier are directly collected from the updated statistical information and the work process. The pass rate data, task execution count, and completion quantity information are then reorganized to obtain an abnormal feature dataset, and abnormal execution features are extracted based on the abnormal feature dataset.

[0079] S3.2: Perform classification and encoding processing on the abnormal execution features in the abnormal feature dataset to obtain the abnormal classification encoding sequence.

[0080] Furthermore, when the risk assessment indicator indicates an early warning, the corresponding early warning node is taken as the processing object. The pass rate data, number of task executions and number of completions of the early warning node are extracted from the abnormal feature dataset to form abnormal execution features. The abnormal execution features are classified and categorized, and the classification information is converted into corresponding abnormal classification codes to obtain an abnormal classification code sequence.

[0081] It should be noted that the classification coding process is for nodes that have triggered warnings. It divides the abnormal execution characteristics into categories according to the differences in characteristics, and assigns a corresponding coding identifier to each category to represent different types of abnormal states.

[0082] The anomaly classification coding sequence is used to compare and classify early warning nodes and attribute faults.

[0083] Mapping processing refers to the corresponding processing of resource identifiers, real-time quality indicator values, and status identifiers in the management multidimensional matrix, and the process of completing status transitions based on the comparison between real-time quality indicator values ​​and performance warning benchmarks. For statuses with a pass rate lower than the performance warning benchmark, differentiated risk level identifiers are generated through fault attribution information mapping. For risk-free statuses, safety level identifiers are generated based on the comparison between real-time quality indicator values ​​and performance warning benchmarks. During the real-time management status interface refresh process, mapping processing also includes color gradient mapping of status identifiers and differentiated risk level identifiers and binding them to interface elements to form corresponding interface display information.

[0084] S3.3: Compare and classify the anomaly classification coding sequence with the anomaly classification coding of each node in the task execution resource pool to obtain fault attribution information. Identify the nodes with matching codes as the same type of fault source and output the attribution information.

[0085] Furthermore, after obtaining the anomaly classification coding sequence, the anomaly classification codes of each node in the task execution resource pool are extracted according to the resource identifier. The anomaly classification codes in the anomaly classification coding sequence are compared and classified with the anomaly classification codes of each node in the task execution resource pool. Nodes with matching codes are classified and organized, and the nodes with matching codes are identified as similar fault sources. Fault attribution information is generated based on the comparison and classification information, and the attribution information (identifying abnormal nodes and similar fault sources) is output.

[0086] S4: Construct a multi-dimensional management matrix based on risk assessment indicators and fault attribution information, including resource, indicator, and status dimensions. Map the risk-free status to a safety level identifier, and generate differentiated risk level identifiers based on the fault attribution information.

[0087] S4.1: The management multidimensional matrix refers to a resource management data structure with the resource identifier as the index dimension and the real-time quality indicator value and status identifier as the data dimension. The status identifier is mapped to the security level identifier and the differentiated risk level identifier based on the comparison between the real-time quality indicator value and the performance warning benchmark, forming a real-time management matrix of resources and pass rate.

[0088] Based on real-time quality metrics, the total task volume and task completion volume corresponding to resource identifiers are reorganized to obtain a multi-dimensional data set corresponding to resources and pass rates.

[0089] Furthermore, based on the real-time quality indicators corresponding to each resource identifier generated by the pass rate calculation, the total task volume and task completion volume corresponding to the resource identifier are extracted and reorganized according to the resource identifier. The resource identifier and the real-time quality indicators are associated and processed. The total task volume, task completion volume and the pass rate corresponding to the real-time quality indicators are organized according to the resource identifier to obtain a multi-dimensional data set corresponding to the resource and the pass rate (the total task volume, task completion volume corresponding to the resource identifier and the real-time quality indicator determined based on the total task volume and task completion volume, i.e., the pass rate).

[0090] S4.2: Based on the multidimensional dataset, perform correlation processing on resource identifiers and real-time quality indicators to obtain a management multidimensional matrix.

[0091] Furthermore, the total task volume, task completion volume, and real-time quality indicators in the multidimensional dataset are associated with resource identifiers. Resource identifiers are used as index dimensions, and real-time quality indicator values ​​and status identifiers are used as data dimensions. Status mapping is performed based on the comparison between real-time quality indicator values ​​and performance warning benchmarks. Each resource identifier corresponds to a real-time quality indicator value and status identifier, forming a real-time management matrix of resources and pass rate, thus obtaining a multidimensional management matrix.

[0092] Specifically, status mapping processing refers to the process of comparing the real-time quality indicator value corresponding to each resource identifier with the performance early warning benchmark based on the resource identifier, real-time quality indicator value and status identifier in the management multidimensional matrix, and then converting the status identifier accordingly based on the comparison between the real-time quality indicator value and the performance early warning benchmark.

[0093] For states where the pass rate is lower than the performance warning benchmark, a differentiated risk level identifier is generated by mapping the fault attribution information. For states without risk, a safety level identifier is generated by mapping the comparison between the real-time quality indicator value and the performance warning benchmark. The resource identifier, real-time quality indicator value, status identifier, safety level identifier, and differentiated risk level identifier are associated to form a real-time management matrix of resources and pass rate that includes the safety level identifier and the differentiated risk level identifier, and the real-time management status interface is refreshed.

[0094] S4.3: By managing a multi-dimensional matrix for state mapping, risk-free states are mapped to safety level identifiers, and risky states are mapped to generate differentiated risk level identifiers based on fault attribution information.

[0095] Furthermore, when performing state mapping processing through the management multidimensional matrix, the real-time quality indicator values ​​corresponding to each resource identifier are compared with the performance early warning benchmark based on the resource identifier, real-time quality indicator values, and state identifier in the management multidimensional matrix, and state mapping processing is performed based on the comparison between the real-time quality indicator values ​​and the performance early warning benchmark.

[0096] For a risk-free state, a safety level identifier is generated by mapping the real-time quality indicator values ​​with the performance warning benchmark. For a state where the pass rate is lower than the performance warning benchmark, a differentiated risk level identifier is generated by mapping the fault attribution information. Resource identifiers, real-time quality indicator values, status identifiers, safety level identifiers, and differentiated risk level identifiers are then associated to form a real-time management matrix of resources and pass rates that includes both safety level identifiers and differentiated risk level identifiers.

[0097] The formula for calculating the differentiated risk level label is:

[0098] ;

[0099] in, Representing the Differentiated risk level identifiers for each resource. Represents a resource identifier index. Representing the The dimensional coefficients for each dimension are obtained based on risk assessment using resource, indicator, and status dimensions. Represents the dimension number. Represents the number of dimensions. Representing the The resource identifier is in the first Each dimension represents a state identifier value.

[0100] It should be noted that the dimensional status values ​​are determined based on fault attribution information, comparison information between real-time quality indicators and performance early warning benchmarks, and risk assessment indicators.

[0101] when When dealing with resource dimensions, the system determines whether the corresponding resource identifier belongs to an abnormal fault source based on the fault attribution information, and generates resource dimension status values. When the indicator dimension is used, the indicator dimension state value is generated based on the degree of deviation of the real-time quality indicator from the performance warning benchmark. When the state dimension is used, the state dimension state value is generated based on the state identifier corresponding to the risk assessment indicator.

[0102] S5: Refresh the real-time management status interface based on the management multi-dimensional matrix, match differentiated risk level indicators and security level indicators, and trigger resource scheduling, task reallocation and anomaly handling processes.

[0103] S5.1: Based on the management multidimensional matrix, data mapping processing is performed, and color gradient mapping is performed on status identifiers and differentiated risk level identifiers and bound to interface elements to obtain the refreshed real-time management status interface.

[0104] Furthermore, when performing data mapping processing based on the management multidimensional matrix, real-time quality indicator values, status indicators, security level indicators, and differentiated risk level indicators are extracted from the management multidimensional matrix according to resource identifiers. The security level indicators corresponding to the risk-free status are mapped with the status indicators using color gradient mapping. The differentiated risk level indicators corresponding to the pass rate being lower than the performance warning benchmark are also mapped with the status indicators using color gradient mapping. The color gradient mapping information is bound to the interface elements corresponding to the resource identifiers. The real-time quality indicator values, status indicators, and differentiated risk level indicators in the interface elements are refreshed according to the update information of the management multidimensional matrix to obtain the refreshed real-time management status interface.

[0105] Specifically, color gradient mapping refers to the process of converting different states into color display levels on the interface based on the status identifiers, differentiated risk level identifiers, and security level identifiers in the management multidimensional matrix, so as to form a visual distinction of corresponding resource identifiers and real-time quality indicator values.

[0106] Interface element binding refers to the process of associating status identifiers, differentiated risk level identifiers, and security level identifiers with the corresponding display content in the real-time management status interface, and synchronously displaying resource identifiers, real-time quality indicator values, and corresponding statuses in the real-time management status interface.

[0107] S5.2: Based on the refreshed real-time management situation interface, combined with differentiated risk level identifiers and security level identifiers, strategy matching is performed to generate resource scheduling and task reallocation control instructions.

[0108] Furthermore, based on the refreshed real-time management situation interface, when performing strategy matching processing on differentiated risk level identifiers and security level identifiers, the real-time quality indicator values, status identifiers, differentiated risk level identifiers, and security level identifiers corresponding to each resource identifier are extracted based on the management multi-dimensional matrix. Based on the comparison between the real-time quality indicator values ​​and the performance warning benchmark, the resource identifiers corresponding to the differentiated risk level identifiers are identified as high-risk resources requiring adjustment, and the resource identifiers corresponding to the security level identifiers are identified as low-risk task-accepting resources. According to the task redistribution control instructions, tasks on high-risk resources requiring adjustment are redistributed to low-risk task-accepting resources. Corresponding matching processing is performed according to resource dimensions, indicator dimensions, and status dimensions to generate resource scheduling and task redistribution control instructions.

[0109] Resource identifiers with pass rates below the performance warning benchmark will have their corresponding task execution relationships adjusted according to task reallocation control instructions. Resource identifiers with no risk status will participate in task execution, thus optimizing resource workflows.

[0110] It should be noted that the strategy matching process is based on a multi-dimensional management matrix. It analyzes the differentiated risk level identifiers, security level identifiers, real-time quality indicator values, and status identifiers corresponding to each resource identifier. Based on the comparison between the real-time quality indicator values ​​and the performance warning benchmark, it classifies and matches resource identifiers with a pass rate lower than the performance warning benchmark and resource identifiers in a risk-free state, and generates resource scheduling and task reallocation control instructions.

[0111] Task reallocation control instructions are control information generated based on differentiated risk level identifiers and security level identifiers, which adjust task execution relationships and allocate tasks to corresponding resource identifiers.

[0112] S5.3: Execute the exception handling process according to the task reassignment control instruction to complete resource scheduling, task reassignment and trigger the exception handling process.

[0113] Furthermore, based on the management multidimensional matrix, the differentiated risk level identifiers and security level identifiers are subjected to policy matching processing. After generating resource scheduling and task redistribution control instructions, the resource identifiers corresponding to the differentiated risk level identifiers and the resource identifiers corresponding to the security level identifiers are associated according to the task redistribution control instructions. This completes the resource scheduling and task redistribution processing within the task execution resource pool. Based on the fault attribution information, the nodes that match the anomaly classification code are identified as similar fault sources. Anomaly handling procedures are executed for similar fault sources, thus triggering the resource scheduling, task redistribution, and anomaly handling procedures.

[0114] This embodiment also provides a real-time pass rate matrix monitoring system, including: a statistics maintenance module, used to receive job records from the task execution resource pool, establish performance statistics data according to resource identifiers, identify and correct data consistency conflicts between data sources through multi-source data fusion, update the total task volume and completion volume, and output real-time quality indicators and updated statistical information; an early warning analysis module, used to extract historical quality indicator sequences based on updated statistical information, calculate statistical feature values ​​for historical quality indicator sequences and generate performance early warning benchmarks, compare real-time quality indicators with performance early warning benchmarks, and obtain risk assessment indicators; a fault clustering module, used to extract abnormal execution features and classify and encode them when the risk assessment indicator indicates an early warning, and generate fault attribution information; a matrix generation module, used to construct a multi-dimensional management matrix with resource, indicator, and status dimensions based on risk assessment indicators and fault attribution information, map risk-free states to safety level identifiers, and generate differentiated risk level identifiers based on fault attribution information; and a visualization and control module, used to refresh the real-time management status interface according to the multi-dimensional management matrix, match differentiated risk level identifiers and safety level identifiers, and trigger resource scheduling, task reallocation, and anomaly handling processes.

[0115] In summary, this invention improves the targeting of scheduling strategies and the accuracy of anomaly handling by constructing a multi-dimensional management matrix, generating differentiated risk level identifiers based on fault attribution information, and triggering resource scheduling based on the differentiated risk level identifiers, thereby ensuring the stable operation of the task execution resource pool.

[0116] 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.

Claims

1. A real-time pass rate matrix monitoring method, characterized in that, include: Receive job records from the task execution resource pool, establish performance statistics according to resource identifiers, identify and correct data consistency conflicts between data sources through multi-source data fusion, update the total task volume and completion volume, and output real-time quality indicators and updated statistical information; Based on the updated statistical information, historical quality indicator sequences are extracted, statistical feature values ​​are calculated for the historical quality indicator sequences, and performance early warning benchmarks are generated. Real-time quality indicators are compared with performance early warning benchmarks to obtain risk assessment indicators. When the risk assessment indicator indicates an early warning, the abnormal execution characteristics are extracted, classified, and coded to generate fault attribution information; Based on risk assessment indicators and fault attribution information, a multi-dimensional management matrix is ​​constructed with resource, indicator, and status dimensions. Risk-free status is mapped to safety level identifiers, and differentiated risk level identifiers are generated based on fault attribution information. The real-time management status interface is refreshed based on the management multi-dimensional matrix, and differentiated risk level and security level indicators are matched to trigger resource scheduling, task reallocation and anomaly handling processes.

2. The real-time pass rate matrix monitoring method as described in claim 1, characterized in that, The task execution resource pool refers to a resource cluster composed of multiple nodes with independent network location information, and each node corresponds to a unique resource identifier for performance tracking. The resource identifier includes network location information, and the job record includes the total number of business tasks generated by the node in parallel processing within the management and statistical period and the number of tasks completed.

3. The real-time pass rate matrix monitoring method as described in claim 2, characterized in that, The specific steps for outputting real-time quality indicators and updated statistical information are as follows: Based on the job records of the task execution resource pool, data from each node is extracted by resource identifier and classified and organized for performance management to obtain performance statistics data distinguished by resource identifier; Extract multi-source business information from performance statistics, identify and merge data consistency conflicts, and obtain updated total task volume and completion statistics. Using the updated total task volume and completion statistics, calculate the real-time quality indicators corresponding to each resource identifier to obtain the updated statistics.

4. The real-time pass rate matrix monitoring method as described in claim 3, characterized in that, The historical quality indicator sequence refers to a performance trend analysis set composed of real-time quality indicators corresponding to each resource identifier within multiple consecutive statistical periods in the past, arranged chronologically according to the management time cycle. The real-time quality indicator is the ratio of the number of tasks completed corresponding to each resource identifier to the total number of tasks, i.e., the pass rate.

5. The real-time pass rate matrix monitoring method as described in claim 4, characterized in that, The specific steps for obtaining the risk assessment indicator are as follows: Based on the updated statistical information, the management time window extraction method is used to collect historical quality indicator data and obtain a historical quality indicator sequence arranged in chronological order. Statistical characteristics are calculated on historical quality indicator sequences to obtain the mean and fluctuation range. The warning interval is determined based on the mean and fluctuation range and used as the performance warning benchmark. Based on the performance early warning benchmark, risk assessment indicators are obtained by comparing and analyzing the differences between real-time quality indicators and the performance early warning benchmark.

6. The real-time pass rate matrix monitoring method as described in claim 5, characterized in that, The specific steps for generating fault attribution information are as follows: Based on risk assessment indicators, abnormal execution features are extracted. By collecting pass rate data, task execution times, and completion quantities during the work process and reorganizing them, an abnormal feature dataset is obtained. The abnormal execution features in the abnormal feature dataset are classified and encoded to obtain an abnormal classification and encoding sequence. By comparing and classifying the anomaly classification coding sequence with the anomaly classification coding of each node in the task execution resource pool, fault attribution information is obtained. Nodes with matching codes are identified as fault sources of the same type and the attribution information is output.

7. The real-time pass rate matrix monitoring method as described in claim 6, characterized in that, The management multidimensional matrix refers to a resource management data structure with the resource identifier as the index dimension and the real-time quality indicator value and status identifier as the data dimension. The status identifier is mapped to a security level identifier and a differentiated risk level identifier based on the comparison between the real-time quality indicator value and the performance warning benchmark, forming a real-time management matrix of resources and pass rate.

8. The real-time pass rate matrix monitoring method as described in claim 7, characterized in that, The specific steps for generating differentiated risk level identifiers are as follows: Based on real-time quality indicators, the total task volume and task completion volume corresponding to resource identifiers are reorganized to obtain a multi-dimensional data set corresponding to resources and pass rates. Based on the multidimensional dataset, resource identifiers and real-time quality indicators are correlated to obtain a multidimensional management matrix. By managing a multi-dimensional matrix for state mapping, risk-free states are mapped to safety level identifiers, and risky states are mapped to generate differentiated risk level identifiers based on fault attribution information.

9. The real-time pass rate matrix monitoring method as described in claim 8, characterized in that, The specific steps of triggering resource scheduling, task reallocation, and exception handling are as follows: Based on the management multidimensional matrix, data mapping processing is performed, and color gradient mapping is applied to status identifiers and differentiated risk level identifiers and bound to interface elements to obtain a refreshed real-time management status interface. Based on the refreshed real-time management situation interface, combined with differentiated risk level identifiers and security level identifiers, strategy matching is performed to generate resource scheduling and task reallocation control instructions. The exception handling process is executed according to the task reassignment control instructions, which completes resource scheduling, task reassignment and triggering of exception handling process.

10. A real-time pass rate matrix monitoring system, based on the real-time pass rate matrix monitoring method according to any one of claims 1 to 9, characterized in that, include: The statistics and maintenance module is used to receive job records from the task execution resource pool, establish performance statistics data according to resource identifiers, identify and correct data consistency conflicts between data sources through multi-source data fusion, update the total task volume and completion volume, and output real-time quality indicators and updated statistical information. The early warning analysis module is used to extract historical quality indicator sequences based on updated statistical information, calculate statistical feature values ​​for the historical quality indicator sequences and generate performance early warning benchmarks, compare real-time quality indicators with performance early warning benchmarks, and obtain risk assessment indicators. The fault clustering module is used to extract abnormal execution features and classify and encode them to generate fault attribution information when the risk assessment indicator indicates an early warning. The matrix generation module is used to construct a multi-dimensional management matrix with resource, indicator, and status dimensions based on risk assessment indicators and fault attribution information. It maps risk-free status to safety level identifiers and generates differentiated risk level identifiers based on fault attribution information. The visualization and control module is used to refresh the real-time management status interface based on the management multi-dimensional matrix, match differentiated risk level identifiers and security level identifiers, and trigger resource scheduling, task reallocation and anomaly handling processes.

Images