A method and system for monitoring data cycles

Abstract

This application relates to the field of data monitoring, and in particular to a method and system for monitoring data cycles. The method includes: obtaining a set A of data transmission paths for raw data packets within a target time period; dividing A into partitions to obtain partition result B; traversing B, if B... r If the number of included data transmission paths is greater than or equal to the preset target number, then according to B... r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r According to D r and C r Judge B r Is the corresponding data transmission path an abnormal transmission path? If so, obtain B. r The corresponding data transmission path includes abnormal nodes, and abnormal warning information is output. This invention can promptly identify abnormal transmission paths and abnormal transmission nodes, improving the reliability and stability of subsequent data packets throughout their lifecycle.

Description

Technical Field

[0001] This invention relates to the field of data monitoring, and in particular to a method and system for monitoring data cycles. Background Technology

[0002] In today's digital age, data transmission, as a core component of information system operation, is widely used in various scenarios such as industrial control and cloud computing services. As system complexity increases, data needs to go through multiple processing nodes to complete its flow. For example, Chinese patent application number 202111491566.5 discloses a data cycle monitoring method and storage medium, which discloses the multi-level transmission phenomenon of data packets during their life cycle, as well as the splitting and merging process of data packets during their life cycle.

[0003] The stability and timeliness of data transmission directly affect the operational efficiency and reliability of the entire system. Existing technologies often focus on monitoring the transmission status of a single data packet throughout its lifecycle, lacking overall monitoring of the transmission status of multiple data packets with the same transmission path throughout their lifecycle. This results in the inability to identify abnormal transmission paths and abnormal transmission nodes in a timely manner, affecting the reliability and stability of subsequent data packets throughout their lifecycle. Summary of the Invention

[0004] The purpose of this invention is to provide a data cycle monitoring method and system to identify abnormal transmission paths and abnormal transmission nodes in a timely manner, thereby improving the reliability and stability of subsequent data packets during their life cycle.

[0005] According to a first aspect of the present invention, a method for monitoring data cycles is provided, the method comprising the following steps:

[0006] S100, Obtain the data transmission path set A of the original data packets within the target time period; A={A1,A2,…,A2} i ,…,A n}, A i Let A be the data transmission path of the i-th original data packet generated within the target time period. i =(A i,1 A i,2 ,…,A i,j ,…,A i,m A i,j The identifier is the j-th processing node through which the i-th raw data packet passes. The value of j ranges from 1 to m, where m is the number of processing nodes through which the i-th raw data packet passes. The value of i ranges from 1 to n, where n is the number of raw data packets generated within the target time period. The raw data packet is a data packet created by the host and that has reached its lifecycle.

[0007] S200, partition A to obtain partition result B; B={B1,B2,…,B r ,…,B R}, B r The r-th data transmission path subset obtained by partitioning A is defined as follows: r ranges from 1 to R, where R is the number of data transmission path subsets obtained by partitioning A; any B r The data transmission paths are the same, but different B r The data transmission paths involved are different.

[0008] S300, iterate through B, if B r If the number of included data transmission paths is greater than or equal to the preset target number, then according to B... r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r B r The processing time deviation sequence corresponding to any data transmission path includes the processing time deviation of the original data packet corresponding to that data transmission path at each processing node included in that data transmission path.

[0009] S400, according to D r and C r Judge B r Is the corresponding data transmission path an abnormal transmission path? If so, obtain B. r The corresponding data transmission path includes abnormal nodes, and abnormal warning information is output.

[0010] According to a second aspect of the present invention, a data cycle monitoring system is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described data cycle monitoring method.

[0011] Compared with the prior art, the present invention has at least the following beneficial effects:

[0012] This invention achieves the segmentation of data transmission paths for raw data packets within a target time period, aggregating identical data transmission paths and analyzing frequently occurring data transmission paths. This avoids indiscriminate monitoring of massive, scattered paths, improving monitoring efficiency. Based on the processing time deviation sequence corresponding to raw data packets with the same transmission path within the target time period, this invention calculates the average deviation and stability of that path, enabling quantitative evaluation of high-frequency transmission paths. This allows for objective judgment of path anomalies, reducing subjective judgment errors. After identifying abnormal transmission paths, this invention further investigates the processing nodes within those paths. Compared to existing technologies that indiscriminately and periodically investigate all processing nodes across all paths, this invention provides maintenance personnel with clearer investigation targets, reducing system fault diagnosis costs. By promptly outputting anomaly warning information, this invention can identify potential risks in data transmission in advance, facilitating timely fault diagnosis and repair, and ensuring the stability and reliability of subsequent data packet transmission throughout its lifecycle. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0014] Figure 1 A flowchart of a data cycle monitoring method provided in Embodiment 1 of the present invention;

[0015] Figure 2 The method for obtaining B as provided in Embodiment 1 of the present invention r The average deviation D of the corresponding data transmission path r And stability C r A flowchart of the process;

[0016] Figure 3 The te provided in Embodiment 1 of the present invention x r,v A flowchart of the acquisition process;

[0017] Figure 4 This is a flowchart illustrating the process of obtaining abnormal transmission paths and abnormal nodes as provided in Embodiment 1 of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1:

[0020] According to this embodiment, as Figure 1 As shown, a method for monitoring data cycles is provided, the method comprising the following steps:

[0021] S100, Obtain the data transmission path set A of the original data packets within the target time period; A={A1,A2,…,A2} i ,…,A n}, A i Let A be the data transmission path of the i-th original data packet generated within the target time period. i =(A i,1 A i,2 ,…,A i,j ,…,A i,m A i,j The identifier is the j-th processing node through which the i-th raw data packet passes. The value of j ranges from 1 to m, where m is the number of processing nodes through which the i-th raw data packet passes. The value of i ranges from 1 to n, where n is the number of raw data packets generated within the target time period. The raw data packet is a data packet created by the host and that has reached its lifecycle.

[0022] In this embodiment, the target time period refers to the monitored time period, such as the most recent hour or the most recent day. The original data packet refers to a data packet generated by the upper-level device (such as a server or terminal device) that has completed its preset lifecycle (such as transmission completion or processing end). The processing node refers to a hardware or software unit (such as a router, server, virtual machine, or application module) that participates in the data transmission process, and its identity is uniquely identified. The data transmission path refers to the ordered sequence of processing nodes that the data packet passes through from its origin to its destination.

[0023] In this embodiment, for raw data packets created by the upper-level terminal that have reached their lifecycle, the identity identifiers of all processing nodes they passed through within the target time period are collected through methods such as log tracking and node recording. These are then arranged into a path sequence according to the transmission order and finally aggregated into a path set A. This embodiment focuses on data packets that have completed their lifecycle, ensuring that the monitored object is a dataset of the complete transmission process and avoiding interference from invalid data that has not been transmitted.

[0024] S200, partition A to obtain partition result B; B={B1,B2,…,B r ,…,B R}, B r The r-th data transmission path subset obtained by partitioning A is defined as follows: r ranges from 1 to R, where R is the number of data transmission path subsets obtained by partitioning A; any B r The data transmission paths are the same, but different B r The data transmission paths involved are different.

[0025] In this embodiment, clustering is performed on A based on the consistency of the path sequence: paths with completely identical transmission paths (identical node sequences) are grouped into the same subset B. r Different paths are grouped into different subsets to achieve path partitioning.

[0026] As a specific implementation method, a string matching or sequence alignment algorithm is used to compare each path A in A one by one. i The node sequence; if the node sequences of two paths are exactly the same (same length and the same node identifier at corresponding positions), then they are classified into the same B. r Otherwise, it will be classified into different categories (B). r Finally, a subset set B is generated, where each B r Includes the unique path and the number of times that path appears.

[0027] Thus, this embodiment achieves accurate classification of transmission paths.

[0028] S300, iterate through B, if B r If the number of included data transmission paths is greater than or equal to the preset target number, then according to B... r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r B r The processing time deviation sequence corresponding to any data transmission path includes the processing time deviation of the original data packet corresponding to that data transmission path at each processing node included in that data transmission path.

[0029] In this embodiment, if the number of data transmission paths included in a subset of data transmission paths is greater than or equal to a preset target number, the data transmission paths corresponding to that subset are determined to be high-frequency paths. Optionally, the preset target number is an empirical value, for example, a preset target number of 100. It should be understood that high-frequency paths have a greater impact on the performance of the transmission system. This embodiment focuses the monitoring on high-frequency paths, which can improve monitoring efficiency. Optionally, if B rIf the number of data transmission paths included is less than the preset target number, then no analysis will be performed or existing monitoring methods will be used for processing.

[0030] As a specific implementation method, such as Figure 2 As shown, according to B r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r include:

[0031] S310, obtain B r The processing time deviation sequence F corresponding to the included data transmission path r F r ={F r,1 ,F r,2 ,…,F r,v ,…,F r,u}, F r,v For B r The processing time deviation sequence F corresponding to the v-th data transmission path is included. r,v =[e 1 r,v ,e 2 r,v ,…,e x r,v ,…,e y r,v ], e x r,v For B r The processing time deviation corresponding to the x-th processing node of the v-th data transmission path is included, where x ranges from 1 to y, and y is B. r The corresponding data transmission path includes the number of processing nodes; the value of v ranges from 1 to u, where u is B. r The number of data transmission paths included.

[0032] It should be understood that B r The original data packets corresponding to the different data transmission paths are different, B r The processing time deviations of the original data packets corresponding to different data transmission paths at the same processing node along the same data transmission path may be the same or different. r The processing time deviation sequence corresponding to any data transmission path includes the processing time deviation of the original data packet corresponding to that data transmission path at each processing node of that data transmission path.

[0033] In this embodiment, the processing time deviation of any processing node in any data transmission path is the absolute difference between the actual processing time and the reference processing time of that processing node in the data transmission path, which can reflect the degree of deviation of the node's processing time. As a specific implementation, e x r,v =|Te x r,v -te x r,v |,Te x r,v For B r The actual processing time of the x-th processing node included in the v-th data transmission path, te x r,v For B r The reference processing time of the xth processing node included in the vth data transmission path.

[0034] In this embodiment, B is dynamically calculated based on context features. r The reference duration te of the xth processing node included in the vth data transmission path. x r,v By analyzing features such as data packets, node status, and business scenarios, a pre-trained model is matched to output targeted reference values. As a preferred implementation method, such as... Figure 3 As shown, te x r,v The acquisition process includes:

[0035] S311, obtain B r The context features G corresponding to the x-th processing node of the v-th data transmission path are included; the context features include data packet features, node real-time status features, and business scenario attribute features; the data packet features include data packet size and priority; the node real-time status features include CPU utilization, memory usage, and number of concurrent tasks; the business scenario attribute features include time period features and date type.

[0036] As an optional specific implementation, the priority is encoded using integers: such as 1 (highest), 2 (high), 3 (medium), 4 (low), 5 (lowest); the time period features are encoded using integers: such as 1 (early morning), 2 (morning peak), 3 (morning work), 4 (noon), 5 (afternoon work), 6 (evening); the date type is encoded using one-hot encoding: weekdays are [1,0,0], weekends are [0,1,0], and statutory holidays are [0,0,1].

[0037] S312, based on G, select matching models from the model list Q corresponding to the target node; the target node is B. rThe data includes the x-th processing node of the v-th data transmission path; Q includes several records, each record including a model and the corresponding reference context features; each model includes the correspondence between the context features and the processing time reference value.

[0038] In this embodiment, S312 includes: obtaining the similarity of the reference context features included in each record of G and Q; if the maximum similarity is greater than or equal to a preset similarity threshold, then the model corresponding to the maximum similarity is determined as the matching model. Optionally, the preset similarity threshold is an empirical value, for example, the preset similarity threshold is 0.8. Optionally, if the maximum similarity is less than the preset similarity threshold, then from the historical processing records of the target node, the k records with the highest similarity to the current context feature G (k is a preset number, such as 5 records) are selected, and the product of the average actual processing time corresponding to these k records and the preset safety coefficient is taken as te. x r,v Optionally, the preset safety factor can be an empirical value, for example, the preset safety factor is 1.2.

[0039] In this embodiment, G and any reference context feature are vectors with the same dimension. Those skilled in the art know that the process of obtaining the similarity between two vectors is an existing technology, such as using cosine similarity to obtain the similarity between two vectors, which will not be described in detail here.

[0040] In this embodiment, the model list Q corresponding to the target node is a pre-established list, where each record includes a pre-trained processing time prediction model for the target node. Each processing time model corresponds to a specific context scenario (such as a large file + low load + weekday scenario, or a large file + high load + holiday scenario). In any processing time model, features are independent variables, the number of independent variables is the number of features included in the context features, and processing time is the dependent variable. Optionally, the processing time prediction model for any node in any scenario under no-anomaly conditions can be obtained by fitting historical data of any node in any scenario. Those skilled in the art will know that the process of fitting data to obtain a model is prior art and will not be described here.

[0041] S313, Substitute G into the matching model and determine the obtained processing time reference value as te. x r,v .

[0042] Based on S311-S313, the obtained reference duration is dynamically adjusted according to the context features, which improves the adaptability of the scene, avoids the bias and misjudgment caused by fixed thresholds, improves the accuracy of the reference value, and provides a reliable benchmark for subsequent bias calculation.

[0043] S320, obtain D r D r=∑ u v=1 ∑ y x=1 e x r,v / y / u.

[0044] In this embodiment, for B r The average of the processing time deviations of all nodes across all transmission paths can reflect B. r The overall average deviation level of the corresponding transmission path.

[0045] S330, obtain C r C r =∑ u v=1 sim(F r,v ,F r,0 ) / u, F r,0 For F r The corresponding median sequence, F r,0 =(e 1 r,0 ,e 2 r,0 ,…,e x r,0 ,…,e y r,0 ), e x r,0 For {e x r,1 ,e x r,2 ,…,e x r,v ,…,e x r,u The median of}, and sim() is used to obtain the similarity.

[0046] In this embodiment, B is calculated. r The mean similarity between the processing time deviation sequence and the median sequence for all included transmission paths can be used to evaluate B. r The fluctuation of the corresponding transmission path deviation; among them, the higher the similarity, the smaller the fluctuation and the better the stability.

[0047] In this embodiment, based on F r Obtain B from the corresponding median sequence r The stability C of the corresponding data transmission path r (The corresponding number of times to obtain similarity is u), compared to obtaining B r The method in this embodiment includes averaging the similarity of any two data transmission paths (corresponding to u×(u-1) / 2 times of similarity acquisition). This method can take into account C... rWhile maintaining accuracy, significantly reduce the number of similarity judgments and improve the acquisition of C. r Efficiency. Those skilled in the art will understand that any prior art method for obtaining the similarity between two sequences falls within the protection scope of this invention, for example, using cosine similarity to obtain the similarity between two sequences.

[0048] In this embodiment, C r C can quantify the stability of path deviation. r The smaller the value, the greater the deviation fluctuation. r The more unstable the corresponding transmission path.

[0049] S400, according to D r and C r Judge B r Is the corresponding data transmission path an abnormal transmission path? If so, obtain B. r The corresponding data transmission path includes abnormal nodes, and abnormal warning information is output.

[0050] In this embodiment, the abnormal warning information includes the abnormal transmission path and the identity identifier of the abnormal node.

[0051] In this embodiment, based on D r and C r Judge B r Whether the corresponding transmission path is abnormal, as a specific implementation method, such as Figure 4 As shown, S400 includes:

[0052] S410, if D r Greater than the preset time deviation threshold, or C r If the value is less than the preset stability threshold, then determine B. r The corresponding transmission path is an abnormal transmission path, and it enters S420.

[0053] In this embodiment, if D r Greater than the preset time deviation threshold and C r If the value is less than the preset stability threshold, proceed to S420. If D... r Less than or equal to the preset time deviation threshold and C r If the value is greater than or equal to the preset stability threshold, then determine B. r The corresponding transmission path is the normal transmission path. Optionally, the preset time deviation threshold and preset stability threshold are both empirical values.

[0054] S420, determine B r Does each processing node in the corresponding data transmission path meet the first preset condition? If B rIf a processing node in the corresponding data transmission path meets a first preset condition, then the processing node is determined to be a suspected node; the first preset condition includes: the proportion of abnormal transmission paths in the specified transmission path is greater than or equal to a preset proportion threshold; the specified transmission path is a subset of data transmission paths whose number of included data transmission paths is greater than or equal to a preset target number.

[0055] Optional, if B r If a processing node in the corresponding data transmission path does not meet the first preset condition, then the processing node is determined not to be a suspected node; wherein, the preset percentage threshold is an empirical value, for example, the preset percentage threshold is 0.3.

[0056] In this embodiment, the specified transmission path is limited to paths corresponding to a subset whose number of data transmission paths is greater than or equal to a preset target number; these are high-frequency paths, as they occur frequently and have a greater impact on the overall stability of the system. This embodiment obtains the proportion of abnormal paths among the high-frequency paths where the processing node is located. If the proportion of abnormal paths among the high-frequency paths where the processing node is located is large, it indicates that the high-frequency path where the processing node is located is more likely to be an abnormal path, and the processing node is more likely to be a risk point, i.e., an abnormal node. Therefore, this embodiment improves the targeting of processing node judgment through the first preset condition screening, enabling more accurate and faster capture of abnormal nodes.

[0057] S430, determine whether the time deviation entropy of each suspected node is greater than the preset time deviation entropy threshold. If the time deviation entropy of a suspected node is greater than the preset time deviation entropy threshold, then the suspected node is determined to be an abnormal node.

[0058] In this embodiment, the time deviation entropy is used to measure the dispersion of the node processing time deviation. The larger the entropy, the more dispersed the deviation distribution and the more violent the fluctuation. The formula for calculating the entropy is existing technology and will not be repeated here.

[0059] Optionally, if the time deviation entropy of a suspected node is less than or equal to a preset time deviation entropy threshold, then it is determined whether the average time deviation of the suspected node is greater than the preset time deviation threshold. If so, the suspected node is identified as an abnormal node; otherwise, the suspected node is identified as a normal node. Optionally, the preset time deviation threshold is an empirical value.

[0060] Based on S410-S430, this embodiment first determines the abnormal transmission path, and then determines the abnormal nodes in the abnormal transmission path. Based on this dual determination mechanism, this embodiment realizes step-by-step positioning from path to node, realizes rapid and accurate source tracing of anomalies, improves investigation efficiency, and also realizes the identification of abnormal paths and abnormal nodes.

[0061] This embodiment divides the data transmission paths of raw data packets within a target time period, aggregates identical data transmission paths, and analyzes frequently occurring data transmission paths, avoiding indiscriminate monitoring of a massive number of scattered paths and improving monitoring efficiency. Based on the processing time deviation sequence corresponding to raw data packets with the same transmission path within the target time period, this embodiment calculates the average deviation and stability of that same transmission path, achieving a quantitative evaluation of high-frequency transmission paths. This allows for objective judgment of path anomalies, reducing subjective judgment errors. After identifying abnormal transmission paths, this embodiment further investigates the processing nodes within those paths. Compared to the existing method of indiscriminately and periodically investigating all processing nodes across all paths, this embodiment provides maintenance personnel with clearer investigation targets, reducing system fault diagnosis costs. By promptly outputting anomaly warning information, this embodiment can detect potential risks in data transmission in advance, facilitating timely fault diagnosis and repair, and ensuring the stability and reliability of subsequent data packet transmission throughout its lifecycle.

[0062] Example 2:

[0063] This embodiment provides a data cycle monitoring system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps:

[0064] Obtain the set A of data transmission paths for the original data packets within the target time period; A = {A1, A2, ..., A...} i ,…,A n}, A i Let A be the data transmission path of the i-th original data packet generated within the target time period. i =(A i,1 A i,2 ,…,A i,j ,…,A i,m A i,j The identifier is the j-th processing node through which the i-th raw data packet passes. The value of j ranges from 1 to m, where m is the number of processing nodes through which the i-th raw data packet passes. The value of i ranges from 1 to n, where n is the number of raw data packets generated within the target time period. The raw data packet is a data packet created by the host and that has reached its lifecycle.

[0065] Partition A to obtain partition B; B = {B1, B2, ..., B} r ,…,B R}, B r The r-th data transmission path subset obtained by partitioning A is defined as follows: r ranges from 1 to R, where R is the number of data transmission path subsets obtained by partitioning A; any B rThe data transmission paths are the same, but different B r The data transmission paths involved are different.

[0066] Iterate through B, if B r If the number of included data transmission paths is greater than or equal to the preset target number, then according to B... r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r B r The processing time deviation sequence corresponding to any data transmission path includes the processing time deviation of the original data packet corresponding to that data transmission path at each processing node included in that data transmission path.

[0067] According to D r and C r Judge B r Is the corresponding data transmission path an abnormal transmission path? If so, obtain B. r The corresponding data transmission path includes abnormal nodes, and abnormal warning information is output.

[0068] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0069] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of the invention. It should also be understood that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. A method for monitoring data cycles, characterized in that, The method includes the following steps: S100, Obtain the data transmission path set A of the original data packets within the target time period; A={A1,A2,…,A2} i ,…,A n }, A i Let A be the data transmission path of the i-th original data packet generated within the target time period. i =(A i,1 A i,2 ,…,A i,j ,…,A i,m ), A i,j The identifier is the j-th processing node through which the i-th raw data packet passes, where j ranges from 1 to m, and m is the number of processing nodes through which the i-th raw data packet passes; i ranges from 1 to n, and n is the number of raw data packets generated within the target time period; the raw data packet is a data packet created by the host and has reached its lifecycle. S200, partition A to obtain partition result B; B={B1,B2,…,B r ,…,B R }, B r The r-th data transmission path subset obtained by partitioning A is defined as follows: r ranges from 1 to R, where R is the number of data transmission path subsets obtained by partitioning A; any B r The data transmission paths are the same, but different B r The data transmission paths involved are different; S300, iterate through B, if B r If the number of included data transmission paths is greater than or equal to the preset target number, then according to B... r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r B r The processing time deviation sequence corresponding to any data transmission path includes the processing time deviation of the original data packet corresponding to the data transmission path at each processing node included in the data transmission path. S400, according to D r and C r Judge B r Is the corresponding data transmission path an abnormal transmission path? If so, obtain B. r The corresponding data transmission path includes abnormal nodes, and abnormal warning information is output; According to B r The processing time deviation sequence corresponding to the included data transmission path is used to obtain B. r The average deviation D of the corresponding data transmission path r And stability C r include: S310, obtain B r The processing time deviation sequence F corresponding to the included data transmission path r F r ={F r,1 ,F r,2 ,…,F r,v ,…,F r,u }, F r,v For B r The processing time deviation sequence F corresponding to the v-th data transmission path is included. r,v =[e 1 r,v ,e 2 r,v ,…,e x r,v ,…,e y r,v ], e x r,v For B r The processing time deviation corresponding to the x-th processing node of the v-th data transmission path is included, where x ranges from 1 to y, and y is B. r The corresponding data transmission path includes the number of processing nodes; the value of v ranges from 1 to u, where u is B. r The number of data transmission paths included; e x r,v =|Te x r,v -te x r,v |,Te x r,v For B r The actual processing time of the x-th processing node included in the v-th data transmission path, te x r,v For B r The reference processing time of the x-th processing node included in the v-th data transmission path; S320, obtain D r D r =∑ u v=1 ∑ y x=1 e x r,v / y / u; S330, obtain C r C r =∑ u v=1 sim(F r,v ,F r,0 ) / u, F r,0 For F r The corresponding median sequence, F r,0 =(e 1 r,0 ,e 2 r,0 ,…,e x r,0 ,…,e y r,0 ), e x r,0 For {e x r,1 ,e x r,2 ,…,e x r,v ,…,e x r,u The median of}, and sim() is used to obtain the similarity.

2. The data cycle monitoring method according to claim 1, characterized in that, The S400 includes: S410, if D r Greater than the preset time deviation threshold, or C r If the value is less than the preset stability threshold, then determine B. r The corresponding transmission path is an abnormal transmission path, and it enters S420; S420, determine B r Does each processing node in the corresponding data transmission path meet the first preset condition? If B r If a processing node in the corresponding data transmission path meets a first preset condition, then the processing node is determined to be a suspected node. The first preset condition includes: the proportion of abnormal transmission paths in the specified transmission path is greater than or equal to a preset proportion threshold; the specified transmission path is a subset of data transmission paths whose number of included data transmission paths is greater than or equal to a preset target number. S430, determine whether the time deviation entropy of each suspected node is greater than the preset time deviation entropy threshold. If the time deviation entropy of a suspected node is greater than the preset time deviation entropy threshold, then the suspected node is determined to be an abnormal node.

3. The data cycle monitoring method according to claim 1, characterized in that, te x r,v The acquisition process includes: S311, obtain B r The context features G corresponding to the x-th processing node of the v-th data transmission path are included; the context features include data packet features, node real-time status features, and business scenario attribute features; the data packet features include data packet size and priority; the node real-time status features include CPU utilization, memory usage, and number of concurrent tasks; the business scenario attribute features include time period features and date type; S312, based on G, select matching models from the model list Q corresponding to the target node; the target node is B. r Includes the x-th processing node of the v-th data transmission path; Q includes several records, each record including a model and the corresponding reference context features of the model; each model includes the correspondence between the context features and the processing time reference value; S313, Substitute G into the matching model and determine the obtained processing time reference value as te. x r,v .

4. The data cycle monitoring method according to claim 3, characterized in that, S312 includes: obtaining the similarity of the reference context features included in each record of G and Q; if the maximum similarity is greater than or equal to a preset similarity threshold, then the model corresponding to the maximum similarity is determined as the matching model.

5. The data cycle monitoring method according to claim 1, characterized in that, The abnormal warning information includes the abnormal transmission path and the identity identifier of the abnormal node.

6. A data cycle monitoring system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the data cycle monitoring method as described in any one of claims 1 to 5.

Images