Micro-service-oriented inter-container abnormal information tracing method and system

Abstract

The invention relates to the technical field of container abnormal information tracing, and provides a microservice-oriented inter-container abnormal information tracing method and system, and the method comprises the steps: generating a service communication performance data set based on a transmission rate index and an error rate index of a collected inter-container communication request; collecting an optical signal intensity physical quantity of a data transmission link bearing the communication request between the containers, and generating a medium attenuation characteristic sequence; generating a channel load competition state set, and constructing a mutually exclusive correlation feature sequence in combination with the dielectric attenuation feature sequence; when abnormal fluctuation data appears in the service communication performance data set, calculating an exclusive response intensity value according to the exclusive association feature sequence, and comparing the exclusive response intensity value with a preset physical dominant threshold and a preset logic dominant threshold to generate an abnormal information traceability result; according to the method, the problems of accuracy and practicability of the traceability result are solved, and the accuracy and practicability of abnormality diagnosis in a large-scale container arrangement environment are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of container abnormal information tracing, and particularly relates to a micro-service-oriented container abnormal information tracing method and system. BACKGROUND

[0002] At present, communication between containers has become a key link to support stable operation of a system in a micro-service architecture. Since frequent data exchange between micro-services depends on complex network links and shared channel resources, once abnormal conditions such as out-of-order data packets or loss occur, it will often lead to service response delay or even call failure. Especially in a large-scale container orchestration environment, problems such as dynamic changes in transmission paths, unstable physical states of links, and intensified competition for logical channel resources are increasingly prominent, making abnormal tracing a core challenge to ensure system reliability.

[0003] Current research attempts to build an end-to-end service communication quality model to assist abnormal tracing. This scheme mainly relies on network performance monitoring tools to collect indicators such as throughput, delay, and packet loss rate of communication between containers, and combines link layer signal quality feedback mechanisms to identify potential abnormal nodes using statistical learning methods. The existing scheme still has obvious limitations in practical application. On the one hand, its perception of physical link state only stays at the coarse-grained signal quality feedback level, and it is difficult to accurately reflect the influence of microscopic physical factors such as medium attenuation on data transmission; on the other hand, it uses a static clustering method to deal with logical channel resource contention problems, which cannot effectively capture the dynamic changing load competition state, affecting the accuracy and practicality of the tracing result. SUMMARY

[0004] The present application provides a micro-service-oriented container abnormal information tracing method and system to solve the problems in the prior art that it is difficult to accurately reflect the influence of microscopic physical factors such as medium attenuation on data transmission; it cannot effectively capture the dynamic changing load competition state, affecting the accuracy and practicality of the tracing result, and to improve the accuracy and practicality of abnormal diagnosis in a large-scale container orchestration environment.

[0005] In a first aspect, the present application provides a micro-service-oriented container abnormal information tracing method, comprising: generating a service communication performance data set based on the transmission rate indicators and error rate indicators of the collected inter-container communication requests; collecting the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request to generate a medium attenuation feature sequence based on the optical signal intensity physical quantity; analyzing the multiplexing logical structure of the shared channel between service instances to generate a channel load competition state set; aligning the medium attenuation feature sequence and the channel load contention state set in time sequence to construct a mutual exclusivity association feature sequence between the transmission medium physical state and the channel logical state; when abnormal fluctuation data appears in the service communication performance data set, according to the mutual exclusivity association feature sequence, an exclusive response intensity value is calculated, the exclusive response intensity value is compared and analyzed with a preset physical dominant threshold and a preset logical dominant threshold, and an abnormal information tracing result is generated, the abnormal information tracing result includes a transmission medium physical abnormality or a logical channel resource contention.

[0006] Optionally, the light signal intensity physical quantity of the data transmission link carrying the inter-container communication request is collected to generate a medium attenuation feature sequence based on the light signal intensity physical quantity, including: Collecting the light signal intensity physical quantity of the data transmission link carrying the inter-container communication request to form a light intensity original sampling sequence; Determine the stable intensity value of the light intensity original sampling sequence in the abnormal-free communication period as a reference baseline; Calculate the difference between the light signal intensity physical quantity of each sampling point and the reference baseline to generate a light signal intensity deviation sequence; Perform directional accumulation processing on the light signal intensity deviation sequence, when the deviation direction of two or more sampling points is detected to be the same, the deviation amounts of the sampling points with the same deviation direction are accumulated as a one-way attenuation amount, and when the deviation direction of the next sampling point changes relative to the previous sampling point, the sum of the deviation amounts accumulated before the change in the deviation direction is recorded as a one-way attenuation amount; According to the sampling time sequence, arrange all one-way attenuation amounts to generate a medium attenuation feature sequence.

[0007] Optionally, the multiplexing logical structure of the shared channel between service instances is analyzed to generate a channel load contention state set, including: Identify the transmission sub-channels in the multiplexing logical structure of the shared channel between service instances, and define a basic time unit corresponding to the minimum unit of data transmission; According to the data transmission situation of each transmission sub-channel in the corresponding basic time unit, mark the data transmission state of each transmission sub-channel in each basic time unit, the data transmission state includes occupied state and idle state; Detect the data transmission states of all transmission sub-channels in the same basic time unit, when the data transmission requests of multiple service instances point to the same transmission sub-channel, record the resource contention event of the corresponding transmission sub-channel, and count the number of resource contention events in each basic time unit; Aggregate the number of resource contention events in all basic time units to form a channel load contention state set.

[0008] Optionally, the data transmission states of all transmission sub-channels in the same basic time unit are detected, and when the data transmission requests of multiple service instances are directed to the same transmission sub-channel, a resource contention event is recorded for the corresponding transmission sub-channel, and the number of resource contention events in each basic time unit is counted, including: Obtaining the data transmission requests issued by all service instances in each basic time unit, and marking the request allocation time stamp of each data transmission request; Detecting the data transmission states of all transmission sub-channels in the same basic time unit, and generating a channel occupation time period sequence for each transmission sub-channel in combination with the corresponding request allocation time stamp; Marking the transmission sub-channel in the idle state as the occupied state when the data transmission request is directed to the target transmission sub-channel marked as the occupied state, and detecting whether the corresponding request allocation time stamp overlaps with the existing channel occupation time period, if the corresponding request allocation time stamp overlaps with the existing channel occupation time period, and there are two or more data transmission requests of service instances in the overlapping period, then a resource contention event is recorded for the target transmission sub-channel; Traverse all transmission sub-channels, and take the number of resource contention events in each basic time unit as the number of resource contention events.

[0009] Optionally, the data transmission states of all transmission sub-channels in the same basic time unit are detected, and a channel occupation time period sequence for each transmission sub-channel is generated in combination with the corresponding request allocation time stamp, including: Arranging all data transmission requests according to the order of all request allocation time stamps, and sequentially processing each data transmission request, if the target transmission sub-channel is in the idle state, then the channel occupation start time is set as the request allocation time stamp of the target transmission sub-channel, and the channel occupation end time is set as the sum of the request allocation time stamp of the target transmission sub-channel and the preset transmission duration, to generate the channel occupation time period of each transmission sub-channel; If the target transmission sub-channel is in the occupied state, then the end time of the last channel occupation time period that already exists in the target transmission sub-channel is set as the channel occupation end time, and when the request allocation time stamp of the target transmission sub-channel is greater than or equal to the corresponding channel occupation end time, then the channel occupation start time is set as the request allocation time stamp of the target transmission sub-channel, and the channel occupation end time is set as the sum of the request allocation time stamp of the target transmission sub-channel and the preset transmission duration, to generate the channel occupation time period of each transmission sub-channel; According to the transmission sub-channel grouping, all channel occupation time periods are integrated to form a channel occupation time period sequence.

[0010] Optionally, the medium attenuation feature sequence and the channel load contention state set are time-aligned to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state, including: The medium attenuation feature sequence is segmented by a basic time unit to obtain a medium attenuation feature value corresponding to each basic time unit, and a maximum unidirectional change amount in all medium attenuation feature values is taken as a physical state representation value; The channel load contention state set is segmented by a basic time unit to obtain a channel load contention value corresponding to each basic time unit, and a number of times that the channel load contention value is continuously greater than a contention threshold in each basic time unit is taken as a logical state representation value; A basic time unit in which more than three consecutive physical state representation values are in a sustained increasing state and a fluctuation amplitude of the logical state representation value is less than a first tolerance is marked as a physical dominant unit; A basic time unit in which the logical state representation value is in a stepped increase and a change amplitude of the physical state representation value is less than a second tolerance is marked as a logical dominant unit; All physical dominant units and logical dominant units are integrated to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state.

[0011] Optionally, when abnormal fluctuation data appears in the service communication performance data set, an exclusive response intensity value is calculated according to the mutual exclusivity association feature sequence, the exclusive response intensity value is compared with a preset physical dominant threshold and a preset logical dominant threshold, and an abnormal information tracing result is generated, the abnormal information tracing result including a transmission medium physical abnormality or a logical channel resource contention, including: An abnormal fluctuation data corresponding time interval in the service communication performance data set is located, and the time interval is taken as a target analysis window; A first frequency of a physical dominant unit and a second frequency of a logical dominant unit in the mutual exclusivity association feature sequence in the target analysis window are extracted, and a ratio of the first frequency to the second frequency is taken as an exclusive response intensity value; When the exclusive response intensity value is greater than the preset physical dominant threshold, it is determined that the medium attenuation feature sequence has a sustained unidirectional change feature in the target analysis window, and a transmission medium physical abnormality tracing result is generated; When the exclusive response intensity value is less than the preset logical dominant threshold, it is determined that the channel load contention state set has a stepped mutation feature in the target analysis window, and a logical channel resource contention tracing result is generated; When the exclusive response intensity value is between the physical dominant threshold and the logical dominant threshold, if the medium attenuation feature sequence meets the one-way growth condition within the target analysis window, a transmission medium physical anomaly is generated, and if the channel load competition state set meets the step mutation condition within the target analysis window, a logical channel resource contention is generated.

[0012] In a second aspect, the present application provides a microservice-oriented inter-container abnormal information tracing system, comprising: A generation module is configured to generate a service communication performance data set based on the transmission rate indicator and the error rate indicator of the collected inter-container communication request; A collection module is configured to collect an optical signal intensity physical quantity of a data transmission link carrying the inter-container communication request, and generate a medium attenuation feature sequence based on the optical signal intensity physical quantity; An analysis module is configured to analyze a multiplexing logical structure of a shared channel between service instances to generate a channel load competition state set; A construction module is configured to time-align the medium attenuation feature sequence and the channel load competition state set to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state; An analysis module is configured to calculate an exclusive response intensity value according to the mutual exclusivity association feature sequence when abnormal fluctuation data appears in the service communication performance data set, compare the exclusive response intensity value with a preset physical dominant threshold and a preset logical dominant threshold, and generate an abnormal information tracing result, wherein the abnormal information tracing result includes a transmission medium physical anomaly or a logical channel resource contention.

[0013] In a third aspect, the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the microservice-oriented inter-container abnormal information tracing method of the first aspect.

[0014] In a fourth aspect, the present application provides a computer storage medium storing a computer program, wherein the computer program is executed by a computer to implement the microservice-oriented inter-container abnormal information tracing method of the first aspect.

[0015] In the application, based on the transmission rate index and error rate index of the collected inter-container communication request, a service communication performance data set is generated; the optical signal strength physical quantity of the data transmission link carrying the inter-container communication request is collected to generate a medium attenuation feature sequence based on the optical signal strength physical quantity; the multiplexing logical structure of the shared channel between service instances is analyzed to generate a channel load competition state set; the medium attenuation feature sequence and the channel load competition state set are time-aligned to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state; when abnormal fluctuation data appears in the service communication performance data set, an exclusive response intensity value is calculated according to the mutual exclusivity association feature sequence, the exclusive response intensity value is compared and analyzed with a preset physical dominant threshold and a preset logical dominant threshold to generate an abnormal information tracing result, and the abnormal information tracing result includes a transmission medium physical abnormality or a logical channel resource contention. The technical scheme provided by the application breaks through the response delay problem caused by the dependence of the prior art on the end-to-end model, realizes real-time perception of micro-service communication abnormalities, and establishes a panoramic monitoring basis for inter-container communication quality by collecting transmission rate indexes and error rate indexes; the directional cumulative processing of the optical signal strength physical quantity realizes the quantification of the microscopic physical characteristic quantity of the optical fiber medium attenuation, overcomes the defect of coarse-grained physical link state perception, and provides a direct evidence chain for distinguishing physical abnormalities; through dynamic competition event statistics in a basic time unit, the burst characteristics of multiplexing channel resource contention are accurately captured, and the core bottleneck that static clustering cannot capture dynamic load changes is solved; a time coupling mechanism of physical attenuation features and logical competition states is established to overcome the defect of unassociated physical / logical states and form the basis for determining mixed abnormal tracing; through threshold decision and feature backtracking of the exclusive response intensity value, the physical medium degradation and logical resource contention are accurately distinguished in a complex abnormal scenario, and the problem of high mixed abnormal misjudgment rate is eliminated. Further, in the micro-service container shared channel scenario, the transmission sub-channels are divided by defining a basic time unit, and the occupation states of each sub-channel are marked in real time; when the data transmission requests of multiple service instances conflict, a resource contention event is triggered, the number of resource contention events in a unit is counted and aggregated to form a channel load competition state set, and dynamic load intensity quantification in a multiplexing environment is realized. The rigid restriction of the existing static clustering method is broken through, the dynamic fluctuation characteristics of channel resource contention are accurately captured through competition event density statistics of micro time units, high-sensitivity quantitative basis is provided for logical layer abnormal tracing, and the defect of being unable to adapt to dynamic load changes is solved.

[0016] These and other aspects of the present application will become more apparent in the following description of embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to make the technical scheme of the present application or the prior art clearer, the accompanying drawings needed in the description of the embodiments or the prior art will be briefly introduced. Obviously, the accompanying drawings described are only some embodiments of the present application, and not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by a person skilled in the art without creative labor fall within the scope of the present application.

[0018] Figure 1 A flowchart of a microservice-oriented inter-container abnormal information tracing method provided by the present application; Figure 2 A structural schematic diagram of a microservice-oriented inter-container abnormal information tracing system provided by the present application; Figure 3 A structural schematic diagram of a computing device provided by the present application. DETAILED DESCRIPTION

[0019] In order to make the technical scheme of the present application or the prior art clearer, the accompanying drawings needed in the description of the embodiments or the prior art will be briefly introduced. Obviously, the accompanying drawings described are only some embodiments of the present application, and not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by a person skilled in the art without creative labor fall within the scope of the present application.

[0020] In some processes described in the specification and claims of the present application and the above-mentioned accompanying drawings, a plurality of operations appearing in a specific order are included, but it should be clearly understood that these operations can be executed or in parallel without the order in which they appear in this text. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes can include more or fewer operations, and these operations can be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this text are used to distinguish different messages, devices, modules, etc., and do not represent the order of precedence. Also, "first" and "second" are not of different types.

[0021] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by a person skilled in the art without creative labor fall within the scope of the present application.

[0022] In view of the problem that it is difficult to accurately trace the inter-container communication exception in the micro-service architecture, especially in the data packet disorder or loss scenario, the existing technology mainly relies on macro network performance indicators such as throughput, delay and packet loss rate, and combines coarse-grained link signal quality feedback and static traffic clustering strategy to locate the exception, which is difficult to effectively cope with the complex coupling relationship between the microstate changes of the physical transmission medium (such as optical signal attenuation) and the dynamic load competition of the logical channel. The existing scheme can only obtain general signal quality information at the physical layer, and lacks the ability of fine-grained collection and analysis of key physical quantities such as optical signal strength; at the logical layer, the static clustering method is used to evaluate the channel load, which cannot accurately reflect the dynamic change process of resource competition under the multiplexing structure, so that the physical exception and logical interference are difficult to distinguish. Therefore, the present application proposes an inter-container exception information tracing method for micro-service, which synchronously collects the performance indicators of inter-container communication, the physical quantities of optical signal strength and the load competition state under the multiplexing structure of the channel, constructs the time-aligned physical-logical correlation feature sequence, and based on the comparison analysis of the exclusive response strength and the preset threshold, accurately identifies the root cause of the exception, thereby effectively solving the defects of the existing technology in the aspects of insufficient distinction between physical and logical exceptions, dynamic response lag and the like, and improving the stability and maintainability of the micro-service system in the complex network environment.

[0023] Figure 1 A flowchart of an inter-container exception information tracing method for micro-service is provided for the embodiments of the present application, as shown in Figure 1 The method comprises the following steps: Step 101: Based on the transmission rate indicators and error rate indicators of the collected inter-container communication requests, a service communication performance data set is generated; In this step, the inter-container communication request refers to the remote procedure call or message passing behavior initiated between the containerized service instances in the micro-service architecture, including the request content, target service identification and priority parameter. The transmission rate indicator refers to the quantitative value reflecting the effective data throughput capacity of the inter-container communication link, which is calculated by dividing the number of successfully transmitted data packets per unit time by the total time. The error rate indicator refers to the measurement value describing the data transmission reliability, which is calculated by dividing the number of transmission failures by the total number of transmissions, including packet loss, check error and timeout exception. The service communication performance data set refers to the time series data set constructed based on the transmission rate and error rate indicators, which is used to represent the micro-service communication quality state, including three-dimensional data of timestamp, indicator value and associated service identification.

[0024] In the embodiment of the present application, by monitoring the inter-container communication process of the micro service, the transmission rate index (the amount of successfully transmitted data in a unit of time) and the error rate index (the frequency of data transmission failure or check error) of the inter-container communication request are collected in real time; based on the preset data aggregation period, the transmission rate index and the error rate index are integrated into a structured data set according to the time window to generate a service communication performance data set.

[0025] Step 102: Collect the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request, to generate a medium attenuation feature sequence based on the optical signal intensity physical quantity; In the embodiment of the present application, the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request is collected to form an optical intensity original sampling sequence, the stable intensity value in the abnormal communication period is determined as a reference baseline, and the difference between the optical signal intensity physical quantity of each sampling point and the reference baseline is calculated to generate an optical signal intensity deviation sequence. The sequence is subjected to directional accumulation processing to obtain a plurality of one-way attenuation amounts, and all one-way attenuation amounts are arranged into a medium attenuation feature sequence according to the sampling time sequence.

[0026] Step 103: Analyze the multiplexing logic structure of the shared channel between service instances to generate a channel load competition state set; In the embodiment of the present application, the transmission subchannel in the multiplexing logic structure of the shared channel between service instances is identified, and a basic time unit corresponding to the minimum unit of data transmission is defined. According to the data transmission of each transmission subchannel in the corresponding basic time unit, the data transmission state of each transmission subchannel in each basic time unit is marked, and the data transmission states of all transmission subchannels in the same basic time unit are detected. When the data transmission request of multiple service instances points to the same transmission subchannel, the resource contention event of the corresponding transmission subchannel is recorded, and the number of resource contention events in each basic time unit is counted. The number of resource contention events in all basic time units is aggregated to form a channel load competition state set.

[0027] Step 104: Time sequence alignment of the medium attenuation feature sequence and the channel load competition state set is performed to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logic state; In the embodiment of the present application, the medium attenuation characteristic sequence is segmented according to the basic time unit, the medium attenuation characteristic value corresponding to each basic time unit is obtained, the maximum one-way change value in all medium attenuation characteristic values is taken as the physical state representation value, the channel load competition state set is segmented according to the basic time unit, the channel load competition value corresponding to each basic time unit is obtained, and the number of times that the channel load competition value in each basic time unit is continuously greater than the competition threshold value is taken as the logic state representation value, the basic time unit in which the physical state representation value is continuously increased for more than three times and the logic state representation value has a fluctuation amplitude less than a first tolerance is marked as a physical dominant unit, the basic time unit in which the logic state representation value is in a ladder type and the physical state representation value has a change amplitude less than a second tolerance is marked as a logic dominant unit, and all the physical dominant units and the logic dominant units are integrated to construct the mutual exclusivity correlation characteristic sequence of the transmission medium physical state and the channel logic state.

[0028] In step 105, when abnormal fluctuation data appears in the service communication performance data set, an exclusive response intensity value is calculated according to the mutual exclusivity correlation characteristic sequence, the exclusive response intensity value is compared with a preset physical dominant threshold value and a preset logic dominant threshold value, and an abnormal information tracing result is generated, the abnormal information tracing result including a transmission medium physical abnormality or a logic channel resource contention. In the embodiment of the present application, the time interval in which the abnormal fluctuation data in the positioning service communication performance data set is located is taken as a target analysis window, the occurrence frequency of the physical dominant unit and the occurrence frequency of the logic dominant unit in the window are extracted, the ratio of the two is calculated as an exclusive response intensity value, when the value is greater than the physical dominant threshold value, it is determined that there is a continuous one-way change characteristic, and a transmission medium physical abnormality result is generated, when the value is less than the logic dominant threshold value, it is determined that there is a ladder type mutation characteristic, and a logic channel resource contention result is generated, and the intermediate value is determined according to the original characteristic condition to trace back the abnormal type.

[0029] In the embodiment of the present application, the medium attenuation characteristic sequence is generated through the directional accumulation processing of the optical signal intensity, the quantitative representation of the microscopic deterioration of the optical fiber medium is realized for the first time, the channel load competition state set is constructed based on the resource contention event statistics of the basic time unit, the sudden conflict in the multiplexing scene is accurately captured, the problem of high mixed abnormality misjudgment rate is solved through the mutual exclusivity correlation characteristic sequence of the physical / logic state, and the accuracy of the container communication abnormality tracing is improved.

[0030] In the embodiment of the present application, the optical signal intensity physical quantity of the data transmission link carrying the container-to-container communication request is collected, and a medium attenuation characteristic sequence is generated based on the optical signal intensity physical quantity, specifically including the following steps: Step 201: collect the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request to form an optical intensity original sampling sequence; In this step, the data transmission link refers to the optical fiber physical channel relied on by the inter-container communication of the microservice, including the signal transmission path composed of the optical cable, the optical module and the photoelectric conversion interface. The optical signal intensity physical quantity refers to the measured physical value reflecting the optical power in the optical fiber, with the unit of decibel milliwatt (dBm), which is directly obtained by the optical power meter. The optical intensity original sampling sequence refers to the time sequence data set composed of the optical signal intensity physical quantity collected at continuous time points, containing the timestamp and intensity value binary tuple.

[0031] In the embodiment of the application, the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request is collected by the photoelectric sensor in real time, and the optical intensity data points are continuously obtained at a fixed sampling frequency. The original optical signal is smoothed by using the sliding window mean filtering technology to eliminate the instantaneous noise interference, and the optical intensity original sampling sequence arranged in time sequence is formed.

[0032] Step 202: determine the stable intensity value of the optical intensity original sampling sequence in the abnormal communication period as the reference baseline; In this step, the abnormal communication period refers to the continuous time period in which the transmission rate index of the inter-container communication is continuously higher than the lower threshold and the error rate index is lower than the upper threshold. The stable intensity value refers to the arithmetic mean value of the optical intensity original sampling sequence in the abnormal communication period, reflecting the reference state of the physical link. The reference baseline refers to the reference value for quantifying signal attenuation, which directly uses the stable intensity value.

[0033] In the embodiment of the application, the history data of the optical intensity original sampling sequence is analyzed to identify the abnormal communication period. The arithmetic mean value of the intensity values of all sampling points in the period is calculated as the stable intensity value, and the stable intensity value is set as the reference baseline.

[0034] Step 203: calculate the difference value between the optical signal intensity physical quantity of each sampling point and the reference baseline to generate the optical signal intensity deviation sequence; In this step, the optical signal intensity deviation sequence refers to the difference sequence composed of the optical signal intensity physical quantity of each sampling point minus the reference baseline, representing the instantaneous signal fluctuation.

[0035] In the embodiment of the application, the baseline offset calculation is performed on each sampling point of the optical intensity original sampling sequence: the optical signal intensity physical quantity of the sampling point is subtracted from the reference baseline to obtain the instantaneous deviation of the sampling point. All instantaneous deviations are integrated in the order of sampling time to generate the optical signal intensity deviation sequence.

[0036] Step 204: performing directional accumulation processing on the light signal intensity deviation sequence, when the deviation direction of two or more sampling points is detected to be the same, the deviation of the sampling points with the same deviation direction is accumulated as a one-way attenuation, and when the deviation direction of the next sampling point is detected to change relative to the previous sampling point, the sum of the deviations accumulated before the change of the deviation direction is recorded as a one-way attenuation; In this step, the deviation direction refers to the mathematical sign of the light signal intensity deviation (a positive sign indicates intensity enhancement, and a negative sign indicates intensity attenuation). The one-way attenuation refers to the sum of the deviations in the same direction, reflecting the cumulative effect of the one-way continuous degradation of the optical fiber medium. The sampling timing refers to the collection time sequence of the light signal intensity physical quantity.

[0037] In the embodiment of the application, the directional accumulation processing is performed on the light signal intensity deviation sequence, specifically: initializing the accumulator to zero, and sequentially scanning each deviation; if the sign (positive / negative) of the current deviation is the same as that of the previous sampling point, the current deviation is accumulated to the accumulator; if the sign changes, the current value of the accumulator is output as a one-way attenuation and the accumulator is reset; finally, all one-way attenuations are output.

[0038] Step 205: arranging all one-way attenuations according to the sampling timing to generate a medium attenuation feature sequence; In this step, the medium attenuation feature sequence refers to a set of one-way attenuation data arranged according to the sampling timing, which is used to describe the dynamic process of physical medium attenuation.

[0039] In the embodiment of the application, all one-way attenuations are arranged according to the sampling timing (i.e. the time sequence of data collection) to form a medium attenuation feature sequence representing the degree of continuous degradation of the optical fiber medium.

[0040] The embodiment of the application converts micro light signal fluctuations into one-way attenuations through directional accumulation processing, breaking through the limitation of the prior art that can only provide static signal quality feedback; the medium attenuation feature sequence realizes the quantitative representation of the progressive degradation of the optical fiber medium, improving the precision of physical layer anomaly detection; combined with dynamic determination during the period of no anomaly communication, the interference of environmental noise on the reference baseline is eliminated, providing reliable physical layer monitoring for micro-service container communication.

[0041] The application provides a specific embodiment, step 103, analyzing the multiplexing logic structure of the shared channel between service instances to generate a set of channel load competition states, specifically including the following steps: Step 301: identifying the transmission subchannel in the multiplexing logic structure of the shared channel between service instances, and defining a basic time unit corresponding to the minimum unit of data transmission; In this step, the service instance refers to a containerized process unit running independently in a microservice architecture. The multiplexing logical structure of the shared channel refers to a physical channel segmentation mechanism shared by multiple service instances, including time slot division of time division multiplexing or frequency band allocation rule of frequency division multiplexing. The transmission subchannel refers to a parallel minimum communication unit in the multiplexing structure, such as a time slot of time division multiplexing or a frequency band of frequency division multiplexing. The minimum data packet size required for a single transmission of inter-container communication refers to the transmission time of the minimum data packet size. The basic time unit refers to a time segment divided based on the length of the minimum data packet size, which is used as a time window for load statistics.

[0042] In the embodiment of the application, by analyzing the network configuration information of the microservice cluster, the parallel transmission subchannels in the multiplexing logical structure of the shared channel between service instances are identified; based on the minimum data packet size of inter-container communication (i.e. the time consumption of single data packet transmission), the time axis is divided into equal-length basic time units, and the length of each unit is an integer multiple of the minimum transmission time consumption.

[0043] Step 302: According to the data transmission situation of each transmission subchannel in the corresponding basic time unit, the data transmission state of each transmission subchannel in each basic time unit is marked, and the data transmission state includes an occupied state and an idle state. In this step, the data transmission situation refers to a binary decision result of whether there is effective data transmission of the transmission subchannel in the basic time unit. The data transmission state refers to the occupied / idle marking of the transmission subchannel, which is generated based on the data transmission situation. The occupied state refers to the current transmission of data of the transmission subchannel. The idle state refers to the current non-transmission of data of the transmission subchannel.

[0044] In the embodiment of the application, the data transmission behavior of each transmission subchannel in each basic time unit is monitored: if the transmission subchannel has data transmission in the unit, it is marked as an occupied state; if there is no data transmission, it is marked as an idle state; and a data transmission state table containing the state markings of all subchannels is generated.

[0045] Step 303: Detecting the data transmission state of all transmission subchannels in the same basic time unit, when the data transmission request of multiple service instances points to the same transmission subchannel, recording the corresponding transmission subchannel resource contention event, and counting the number of resource contention events in each basic time unit. In this step, the resource contention event refers to a conflict event caused by the request of multiple service instances to the same transmission subchannel at the same time.

[0046] In the embodiment of the present application, the data transmission requests sent by all service instances in each basic time unit are obtained, and the request allocation time stamp of each data transmission request is marked; the data transmission states of all transmission sub-channels in the same basic time unit are detected, and the channel occupation period sequence of each transmission sub-channel is generated in combination with the corresponding request allocation time stamp; the transmission sub-channel to which the data transmission request is directed and which is in an idle state is marked as an occupied state, and the occupation start time and the occupation end time are recorded; when the data transmission request is directed to the target transmission sub-channel which is marked as an occupied state, it is detected whether the corresponding request allocation time stamp overlaps with the existing channel occupation period, if the corresponding request allocation time stamp overlaps with the existing channel occupation period, and there are more than two data transmission requests of service instances in the overlapping period, the resource contention event of the target transmission sub-channel is recorded; the number of resource contention events in each basic time unit is taken as the resource contention event quantity by traversing all transmission sub-channels.

[0047] Step 304: The resource contention event quantities in all basic time units are aggregated to form a channel load competition state set; In this step, the channel load competition state set refers to a sequence of resource contention event quantities arranged in the order of basic time units, reflecting the logical channel competition intensity.

[0048] In the embodiment of the present application, the resource contention event quantities of continuous multiple basic time units are aggregated in time sequence to construct a channel load competition state set reflecting the time sequence change of channel competition intensity, such as [t1: 5, t2: 8, t3: 3] representing the resource contention event quantities of three basic time units.

[0049] The embodiment of the present application realizes millisecond-level load monitoring granularity through basic time units, breaks through the minute-level response delay of traditional static clustering methods; based on real-time statistics of resource contention events, the instantaneous conflict intensity of multiplexing channels is accurately quantified; the channel load competition state set provides high-sensitivity logical layer diagnosis basis for micro-service resource contention exceptions, and reduces the misjudgment rate.

[0050] The present application provides a specific embodiment, step 303, detecting the data transmission states of all transmission sub-channels in the same basic time unit, recording the resource contention event of the corresponding transmission sub-channel when the data transmission requests of multiple service instances are directed to the same transmission sub-channel, and counting the number of resource contention events in each basic time unit, which specifically includes the following steps: Step 311: Obtain the data transmission requests sent by all service instances in each basic time unit, and mark the request allocation time stamp of each data transmission request; In this step, the data transmission request refers to the channel use application initiated by the service instance, containing the target subchannel identification, data size and priority parameters. The request allocation timestamp refers to the accurate time mark (format: year-month-day-hour-minute-second-microsecond) of the request arriving at the channel scheduler, used for timing conflict detection.

[0051] In the embodiment of the application, through the communication agent module of the micro-service cluster, the data transmission requests issued by all service instances in each basic time unit are captured in real time; a request allocation timestamp is added to each request to form a request queue with a timestamp.

[0052] Step 312: Detect the data transmission state of all transmission subchannels in the same basic time unit, and generate a channel occupation period sequence of each transmission subchannel in combination with the corresponding request allocation timestamp. In this step, the channel occupation period sequence refers to a set of transmission subchannel occupation records arranged in time sequence, and each record contains an occupation start time (period start time) and an occupation end time (period end time).

[0053] Step 313: Mark the transmission subchannel in the idle state to which the data transmission request points as the occupied state, and when the data transmission request points to the target transmission subchannel marked as the occupied state, detect whether the corresponding request allocation timestamp overlaps with the existing channel occupation period, and if the corresponding request allocation timestamp overlaps with the existing channel occupation period and there are two or more data transmission requests of service instances in the overlapping period, record that the target transmission subchannel has a resource contention event. In the embodiment of the application, the request pointing to the transmission subchannel in the idle state immediately updates the state of the subchannel to the occupied state, and when another data transmission request points to the target transmission subchannel that has been occupied, detect whether the request allocation timestamp of the data transmission request overlaps with any existing channel occupation period (i.e., the request timestamp is less than the occupation end time); if it overlaps and there are two or more data transmission requests of service instances in the overlapping period, record that the target transmission subchannel has a resource contention event.

[0054] Step 314: Traverse all transmission subchannels, and take the number of resource contention events in each basic time unit as the number of resource contention events. In the embodiment of the application, all transmission subchannels in the current basic time unit are traversed, the total number of recorded resource contention events is accumulated, and the number of resource contention events of the unit is output.

[0055] The embodiment of the application realizes dynamic load capture with microsecond-level precision by modeling channel occupancy driven by request allocation timestamps, breaks through the hysteresis of traditional minute-level statistics, accurately identifies instantaneous resource contention events based on the overlapping period multi-request detection mechanism, improves the accuracy of logical channel conflict detection, provides real-time channel competition portraits for high-concurrency microservice scenarios, and supports container resource elastic scaling decisions.

[0056] The embodiment of the application provides a specific embodiment, step 312, detecting the data transmission state of all transmission subchannels in the same basic time unit, generating a channel occupancy period sequence of each transmission subchannel in combination with the corresponding request allocation timestamp, and specifically comprising the following steps: Step 321: Arranging all data transmission requests according to the order of all request allocation timestamps, and sequentially processing each data transmission request, if the target transmission subchannel is in an idle state, setting the channel occupancy start time as the request allocation timestamp of the target transmission subchannel, setting the channel occupancy end time as the sum of the request allocation timestamp of the target transmission subchannel and the preset transmission duration, to generate the channel occupancy period of each transmission subchannel. In this step, the channel occupancy start time refers to the absolute time point at which the transmission subchannel starts transmitting data, which is directly taken from the request allocation timestamp. The target transmission subchannel refers to a specific subchannel identifier (such as frequency band F3 or time slot S5) pointed to by the current data transmission request. The channel occupancy end time refers to the absolute time point at which the transmission subchannel is released, which is calculated by adding the request allocation timestamp and the preset transmission duration. The preset transmission duration refers to the fixed transmission time consumed according to the minimum data packet size of the container communication and the channel bandwidth. The channel occupancy period refers to the closed interval defined by the channel occupancy start time and the channel occupancy end time, indicating the continuous occupancy state of the subchannel.

[0057] In the embodiment of the application, all data transmission requests are sorted from early to late according to the request allocation timestamp. When sequentially processing each data transmission request, if the target transmission subchannel (i.e., the transmission subchannel pointed to by the data transmission request) is in an idle state, a channel occupancy period is created: the channel occupancy start time is set to the request allocation timestamp of the data transmission request, the channel occupancy end time is set to the sum of the request timestamp and the preset transmission duration, and a new occupancy period of the transmission subchannel is generated.

[0058] Step 322: If the target transmission subchannel is in an occupied state, the end time of the last channel occupation period existing in the target transmission subchannel is set as the channel occupation end time, when the request allocation timestamp of the target transmission subchannel is greater than or equal to the corresponding channel occupation end time, the channel occupation start time is set as the request allocation timestamp corresponding to the target transmission subchannel, and the channel occupation end time is set as the sum of the request allocation timestamp corresponding to the target transmission subchannel and the preset transmission duration, so as to generate the channel occupation period of each transmission subchannel; In this step, the last channel occupation period refers to the record with the latest time in the period sequence existing in the target subchannel.

[0059] In the embodiment of the application, if the target transmission subchannel is in an occupied state, the end time of the last channel occupation period is obtained; when the request allocation timestamp of the current data transmission request is greater than or equal to the end time, a new period is created, the channel occupation start time = the request allocation timestamp, and the channel occupation end time = the request allocation timestamp + the preset transmission duration, otherwise, the data transmission request is skipped (period overlap is not created).

[0060] Step 323: According to the transmission subchannel grouping, all channel occupation periods are integrated to form a channel occupation period sequence. In this step, the transmission subchannel grouping refers to classifying and integrating the channel occupation periods according to the unique identifier of the subchannel (such as the frequency band number / slot number).

[0061] In the embodiment of the application, after the processing of all data transmission requests is completed, the total channel occupation periods of each transmission subchannel are integrated according to the transmission subchannel grouping (such as the ascending order of channel ID) to form a structured channel occupation period sequence, such as {F1: [channel occupation period 1, channel occupation period 2], F2: [channel occupation period 3]}.

[0062] The embodiment of the application realizes the dynamic resource occupation modeling with microsecond-level precision through the channel allocation driven by the request timestamp, avoids the high overhead of the traditional polling detection through the conflict avoidance mechanism based on the end time of the last period, and provides the fine-grained evidence chain for the channel competition analysis through the structured period sequence, so that the real-time performance of the resource contention determination is improved.

[0063] The application provides a specific embodiment, step 104, time sequence alignment of the medium attenuation feature sequence and the channel load competition state set is performed to construct the mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state, and the specific steps include the following steps: Step 401: segment the medium attenuation characteristic sequence by basic time units to obtain a medium attenuation characteristic value corresponding to each basic time unit, and take the maximum one-way change amount in all medium attenuation characteristic values as a physical state representation value; In this step, the medium attenuation characteristic value refers to a single data point of the medium attenuation characteristic sequence in a basic time unit, in decibels (dB). The maximum one-way change amount refers to the minimum negative value (such as -0.8 dB) in all medium attenuation characteristic values in a basic time unit, reflecting the maximum attenuation intensity. The physical state representation value refers to a quantitative value representing the physical medium degradation intensity, which directly takes the maximum one-way change amount.

[0064] In the embodiment of the application, the medium attenuation characteristic sequence is segmented by basic time units; all medium attenuation characteristic values in each unit are scanned, and the negative change value with the largest absolute value is selected as the maximum one-way change amount; and the value is taken as the physical state representation value of the current basic time unit.

[0065] Step 402: segment the channel load competition state set by basic time units to obtain a channel load competition value corresponding to each basic time unit, and count the number of times that the channel load competition value in each basic time unit is greater than the competition threshold value continuously as a logic state representation value; In this step, the channel load competition value refers to the number of resource contention events in a basic time unit. The competition threshold value refers to the minimum number of events that triggers the logical exception determination (such as ≥ 3 times per basic time unit). The logic state representation value refers to a quantitative value of the channel competition intensity, which is obtained by counting the number of times that the channel load competition value exceeds the threshold continuously.

[0066] In the embodiment of the application, the channel load competition state set is segmented by the same basic time unit; for the channel load competition value in each basic time unit, the number of times that it is greater than the competition threshold value continuously is counted, and the number of times is taken as the logic state representation value of the current unit.

[0067] Step 403: mark the basic time unit as a physical dominant unit if more than three consecutive physical state representation values are in a sustained increasing state and the fluctuation range of the logic state representation value is less than a first tolerance; In this step, the first tolerance refers to the maximum deviation allowed for the logic state fluctuation, which is used to exclude interference. The physical dominant unit refers to a basic time unit that is determined to be dominated by the physical medium exception.

[0068] In the embodiment of the application, whether the physical state representation values of more than three consecutive basic time units are in a monotonic increasing state, i.e., the next value is strictly greater than the previous value, and the fluctuation range of the logic state representation value in the same period is less than the first tolerance, is detected; if yes, the basic time unit is marked as a physical dominant unit.

[0069] Step 404: Mark the basic time unit as a logical dominant unit when the logical state representation value is stepped up and the change range of the physical state representation value is less than the second tolerance; In this step, the second tolerance refers to the maximum deviation allowed for the physical state fluctuation. The logical dominant unit refers to the basic time unit mark determined as the logical resource contention dominant.

[0070] In the embodiment of the present application, whether the logical state representation value of the adjacent basic time unit is stepped up, that is, the increment exceeds the preset step threshold, and the change range of the physical state representation value in the same period is less than the second tolerance is detected. If it is satisfied, the basic time unit is marked as a logical dominant unit.

[0071] Step 405: Integrate all physical dominant units and logical dominant units to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state; In this step, the transmission medium physical state refers to the health state of the physical medium such as optical fiber, which is represented by the physical dominant unit. The channel logical state refers to the resource usage state of the multiplexing channel, which is represented by the logical dominant unit. The mutual exclusivity association feature sequence refers to the physical / logical dominant unit mark sequence arranged in time sequence, which is used for abnormal traceability decision.

[0072] In the embodiment of the present application, all marked physical dominant units and logical dominant units are integrated in time sequence to form a mutual exclusivity association feature sequence representing the mutual exclusivity of the transmission medium physical state (optical fiber degradation) and the channel logical state (resource contention).

[0073] The embodiment of the present application quantifies the microscopic degradation of the physical medium by the maximum unidirectional change amount, solves the problem of coarse granularity of physical perception, and dynamically captures the channel contention strength based on the number of continuous exceedances, solves the drawbacks of static clustering failure.

[0074] The present application provides a specific embodiment, step 105, when abnormal fluctuation data appears in the service communication performance data set, according to the mutual exclusivity association feature sequence, the exclusive response strength value is calculated, the exclusive response strength value is compared and analyzed with the preset physical dominant threshold and the preset logical dominant threshold, and the abnormal information traceability result is generated, the abnormal information traceability result includes transmission medium physical abnormality or logical channel resource contention, specifically including the following steps: Step 501: Locate the time interval corresponding to the abnormal fluctuation data in the service communication performance data set, and take the time interval as the target analysis window; In this step, the abnormal fluctuation data refers to the abnormal data point set in the service communication performance data set, in which the transmission rate drops by >30% or the error rate rises by >50%. The target analysis window refers to the diagnosis time interval formed by extending 5 basic time units before and after the abnormal fluctuation period as the center.

[0075] In the embodiment of the present application, the transmission rate and error rate indicators in the scanning service communication performance data set are scanned, and when the transmission rate decreases by more than 30% or the error rate increases by more than 50% in three consecutive sampling periods, it is determined that abnormal fluctuation data occurs; 5 basic time units before and after the fluctuation period are expanded to form a target analysis window.

[0076] Step 502: Extracting the first frequency of the appearance of the physical dominant unit and the second frequency of the appearance of the logical dominant unit in the mutually exclusive correlation feature sequence in the target analysis window, and taking the ratio of the first frequency and the second frequency as the exclusive response strength value; In this step, the first frequency refers to the total number of occurrences of the physical dominant unit in the target analysis window. The second frequency refers to the total number of occurrences of the logical dominant unit in the target analysis window. The exclusive response strength value refers to the ratio of the physical dominant unit frequency to the logical dominant unit frequency (first frequency / second frequency).

[0077] In the embodiment of the present application, all the marks in the mutually exclusive correlation feature sequence in the target analysis window are extracted: the first frequency is obtained by counting the number of occurrences of the physical dominant unit, and the second frequency is obtained by counting the number of occurrences of the logical dominant unit; The quotient obtained by dividing the first frequency by the second frequency is taken as the exclusive response strength value.

[0078] Step 503: When the exclusive response strength value is greater than a preset physical dominant threshold value, it is determined that the medium attenuation feature sequence has a sustained unidirectional change feature in the target analysis window, and a traceability result of a transmission medium physical anomaly is generated; In this step, the preset physical dominant threshold value refers to an empirical value of 2.0, and when the exclusive response strength value is greater than 2.0, it is determined that the physical anomaly is dominant. The sustained unidirectional change feature refers to a continuous decrease of more than 5 units of medium attenuation feature values and a single-step decrease of more than 0.2 dB. The traceability result of the transmission medium physical anomaly refers to a diagnostic conclusion report declaring the failure of physical equipment such as optical fibers and optical modules.

[0079] In the embodiment of the present application, when the exclusive response strength value is greater than the preset physical dominant threshold value (i.e. the empirical value 2.0), it is determined that the medium attenuation feature sequence has a sustained unidirectional change feature in the target window, i.e. the attenuation value monotonically decreases for more than 5 basic time units, and a traceability result of a transmission medium physical anomaly is generated.

[0080] Step 504: When the exclusive response strength value is less than a preset logical dominant threshold value, it is determined that the channel load competition state set has a stepwise mutation feature in the target analysis window, and a traceability result of logical channel resource contention is generated; In this step, the preset logic dominant threshold value is an experience value 0.5, and the logic abnormality is determined when the response intensity is less than 0.5. The ladder mutation feature refers to that the channel load competition value increases by more than 3 in adjacent units and is maintained for more than 3 units. The traceability result of the logic channel resource competition refers to a diagnostic conclusion report declaring that the multiplexing channel resource is overloaded.

[0081] In the embodiment of the application, when the exclusive response intensity value is less than the preset logic dominant threshold value (an experience value 0.5), it is determined that the channel load competition state set has the ladder mutation feature in the target window, that is, the competition value in adjacent units increases by more than 3, and the traceability result of the logic channel resource competition is generated.

[0082] Step 505: When the exclusive response intensity value is between the physical dominant threshold value and the logic dominant threshold value, if the medium attenuation feature sequence meets the one-way growth condition in the target analysis window, the traceability result of the transmission medium physical abnormality is generated, and if the channel load competition state set meets the ladder mutation condition in the target analysis window, the traceability result of the logic channel resource competition is generated. In this step, the one-way growth condition refers to that the medium attenuation values of three consecutive units decrease and the cumulative amplitude is greater than 0.9 dB. The ladder mutation condition refers to that the channel load competition value in adjacent basic time units directly jumps from less than or equal to 2 to greater than or equal to 5.

[0083] In the embodiment of the application, when the exclusive response intensity value is between the physical dominant threshold value and the logic dominant threshold value, it is detected whether the medium attenuation feature sequence meets the one-way growth condition (there are three consecutive units with decreasing attenuation values and an amplitude greater than 0.3 dB), and if it meets, the traceability result of the transmission medium physical abnormality is generated. It is detected whether the channel load competition state set meets the ladder mutation condition (there are adjacent units with competition values jumping from less than or equal to 2 to greater than or equal to 5), and if it meets, the traceability result of the logic channel resource competition is generated.

[0084] The embodiment of the application solves the defect that the traditional scheme cannot distinguish the abnormal type by quantifying the physical / logic abnormality dominant degree through the exclusive response intensity value. The innovative double-threshold decision and feature traceback mechanism can directly determine the abnormal type and confirm the abnormal root.

[0085] Figure 2 A structure diagram of a micro-service-oriented inter-container abnormal information traceability system is provided for the embodiment of the application, as shown in FIG. 1. Figure 2 The system comprises: A generation module 21 is configured to generate a service communication performance data set based on the transmission rate indicator and the error rate indicator of the collected inter-container communication request. An acquisition module 22 is configured to acquire an optical signal intensity physical quantity of a data transmission link carrying the inter-container communication request, and generate a medium attenuation feature sequence based on the optical signal intensity physical quantity. The analysis module 23 is configured to analyze the multiplexing logic structure of the service instance shared channel to generate a channel load competition state set; The construction module 24 is configured to time-align the medium attenuation feature sequence and the channel load competition state set to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logic state; The analysis module 25 is configured to, when abnormal fluctuation data appears in the service communication performance data set, calculate an exclusive response intensity value according to the mutual exclusivity association feature sequence, compare the exclusive response intensity value with a preset physical dominant threshold and a preset logic dominant threshold, and generate an abnormal information tracing result, the abnormal information tracing result including a transmission medium physical abnormality or a logic channel resource contention.

[0086] Figure 2 The microservice-oriented inter-container abnormal information tracing system can perform Figure 1 The microservice-oriented inter-container abnormal information tracing method of the embodiments described above has the implementation principle and technical effects which will not be repeated. The specific operation manner of each module and unit of the microservice-oriented inter-container abnormal information tracing system described above has been described in detail in the embodiments related to the method, and will not be described in detail here.

[0087] In one possible design, Figure 2 The microservice-oriented inter-container abnormal information tracing system of the embodiments described above can be implemented as a computing device, such as a server or a terminal. Figure 3 As shown in the figure, the computing device can include a storage component 31 and a processing component 32. The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32.

[0088] The processing component 32 is configured to perform the above Figure 1 The microservice-oriented inter-container abnormal information tracing method of the embodiments described above.

[0089] The processing component 32 can include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component can also be one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components, for executing the above method.

[0090] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or nonvolatile storage devices, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic or optical disk.

[0091] Of course, the computing device can also necessarily include other components, such as an input / output interface, a display component, a communication component, etc.

[0092] The input / output interface provides an interface between the processing component and peripheral interface modules, which can be output devices, input devices, etc.

[0093] The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, etc.

[0094] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform, and the computing device can be a cloud server, and the processing component, the storage component, etc. can be a basic server resource rented or purchased from the cloud computing platform.

[0095] The embodiment of the application also provides a computer storage medium, which stores a computer program, and the computer program can implement the above-mentioned Figure 1 The embodiment shown in the figure is a microservice-oriented container exception information tracing method.

[0096] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiment, which will not be described here.

[0097] The device embodiment described above is only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme. Those skilled in the art can understand and implement it without creative labor.

[0098] Those skilled in the art can clearly understand the technical solutions of the various embodiments from the above description of the embodiments, and the various embodiments can be implemented by means of software with the necessary general hardware platforms, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that makes a contribution, can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0099] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some technical features therein; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for microservice-oriented inter-container abnormal information tracing, characterized in that, The method comprises the following steps: Based on the transmission rate index and the error rate index of the collected inter-container communication request, a service communication performance data set is generated; Collect the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request, and generate a medium attenuation feature sequence based on the optical signal intensity physical quantity; Analyze the multiplexing logical structure of the shared channel between service instances to generate a channel load competition state set; The medium attenuation feature sequence and the channel load competition state set are time-aligned to construct a mutual exclusivity association feature sequence of the transmission medium physical state and the channel logical state; When abnormal fluctuation data appears in the service communication performance data set, the exclusive response intensity value is calculated according to the mutual exclusivity association feature sequence, and the exclusive response intensity value is compared and analyzed with the preset physical dominant threshold and the preset logical dominant threshold to generate an abnormal information tracing result, which includes transmission medium physical abnormalities or logical channel resource contention.

2. The microservice-oriented inter-container exception information tracing method of claim 1, wherein, Collect the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request, and generate a medium attenuation feature sequence based on the optical signal intensity physical quantity, comprising: Collect the optical signal intensity physical quantity of the data transmission link carrying the inter-container communication request to form an optical intensity original sampling sequence; Determine the stable intensity value of the optical intensity original sampling sequence in the abnormal-free communication period as the reference baseline; Calculate the difference between the optical signal intensity physical quantity of each sampling point and the reference baseline to generate an optical signal intensity deviation sequence; Perform directional accumulation processing on the optical signal intensity deviation sequence, and when the deviation direction of two or more sampling points is detected to be the same, the deviation amounts of the sampling points with the same deviation direction are accumulated as unidirectional attenuation amounts, and when the deviation direction of the next sampling point changes relative to the previous sampling point, the total sum of the deviation amounts accumulated before the change of the deviation direction is recorded as the unidirectional attenuation amount; According to the sampling time sequence, arrange all the unidirectional attenuation amounts to generate a medium attenuation feature sequence.

3. The microservice-oriented inter-container exception information tracing method of claim 1, wherein, Analyze the multiplexing logical structure of the shared channel between service instances to generate a channel load competition state set, comprising: Identify the transmission sub-channels in the multiplexing logical structure of the shared channel between service instances, and define a basic time unit corresponding to the minimum unit of data transmission; According to the data transmission situation of each transmission sub-channel in the corresponding basic time unit, mark the data transmission state of each transmission sub-channel in each basic time unit, and the data transmission state includes the occupied state and the idle state; Detect the data transmission states of all transmission sub-channels in the same basic time unit, and when the data transmission requests of multiple service instances point to the same transmission sub-channel, record the resource contention event of the corresponding transmission sub-channel, and count the number of resource contention events in each basic time unit; Aggregate the number of resource contention events in all basic time units to form a channel load competition state set.

4. The microservice-oriented inter-container exception information tracing method according to claim 3, characterized in that, Detecting data transmission states of all transmission sub-channels in a same basic time unit, recording a resource contention event of a corresponding transmission sub-channel when data transmission requests of multiple service instances are directed to the same transmission sub-channel, and counting a number of resource contention events in each basic time unit, including: Obtaining data transmission requests issued by all service instances in each basic time unit, and marking a request allocation time stamp of each data transmission request; Detecting data transmission states of all transmission sub-channels in a same basic time unit, and generating a channel occupation time period sequence of each transmission sub-channel in combination with a corresponding request allocation time stamp; Marking a transmission sub-channel in an idle state as an occupied state when a data transmission request is directed to the target transmission sub-channel, detecting whether a corresponding request allocation time stamp overlaps with an existing channel occupation time period, and recording a resource contention event of the target transmission sub-channel when the corresponding request allocation time stamp overlaps with the existing channel occupation time period and there are more than two data transmission requests of service instances in the overlapping time period; Counting a number of resource contention events in each basic time unit by traversing all transmission sub-channels.

5. The microservice-oriented inter-container exception information tracing method of claim 4, wherein, Detecting data transmission states of all transmission sub-channels in a same basic time unit, and generating a channel occupation time period sequence of each transmission sub-channel in combination with a corresponding request allocation time stamp, including: Arranging all data transmission requests according to an order of all request allocation time stamps, and sequentially processing each data transmission request, setting a channel occupation start time as a request allocation time stamp of a target transmission sub-channel and setting a channel occupation end time as a sum of the request allocation time stamp of the target transmission sub-channel and a preset transmission duration to generate a channel occupation time period of each transmission sub-channel if the target transmission sub-channel is in an idle state; Setting an end time of a last channel occupation time period of a target transmission sub-channel as the channel occupation end time if the target transmission sub-channel is in an occupied state, setting the channel occupation start time as a request allocation time stamp of the target transmission sub-channel and setting the channel occupation end time as a sum of the request allocation time stamp of the target transmission sub-channel and a preset transmission duration to generate a channel occupation time period of each transmission sub-channel when the request allocation time stamp of the target transmission sub-channel is greater than or equal to a corresponding channel occupation end time; Integrating all channel occupation time periods according to transmission sub-channel groups to form a channel occupation time period sequence.

6. The microservice-oriented inter-container exception information tracing method of claim 1, wherein, Timing aligning the medium attenuation feature sequence and the channel load contention state set to construct a mutual exclusivity association feature sequence of a transmission medium physical state and a channel logical state, including: Segmenting the medium attenuation feature sequence according to basic time units to obtain a medium attenuation feature value corresponding to each basic time unit, and taking a maximum one-way change amount in all medium attenuation feature values as a physical state representation value; segmenting the channel load competition state set by a basic time unit to obtain a channel load competition value corresponding to each basic time unit, and counting a number of times that the channel load competition value in each basic time unit is continuously greater than a competition threshold value as a logical state representation value; a basic time unit in which the three or more physical state representation values are in a sustained increasing state and the logical state representation value fluctuates by a magnitude less than a first tolerance is marked as a physical dominant unit; a basic time unit in which the logical state representation value is in a stepwise increasing state and the change in the physical state representation value has a magnitude less than a second tolerance is marked as a logical dominant unit; integrating all the physical dominant units and the logical dominant units to construct a mutual exclusivity association feature sequence of a transmission medium physical state and a channel logical state.

7. The microservice-oriented inter-container exception information tracing method of claim 1, wherein, When abnormal fluctuation data appears in the service communication performance data set, an exclusive response intensity value is calculated according to the mutual exclusivity association feature sequence, the exclusive response intensity value is compared with a preset physical dominant threshold value and a preset logical dominant threshold value, and an abnormal information tracing result is generated, which includes a transmission medium physical abnormality or a logical channel resource contention, including: locating a time interval corresponding to the abnormal fluctuation data in the service communication performance data set, and taking the time interval as a target analysis window; extracting a first frequency of the physical dominant unit and a second frequency of the logical dominant unit in the mutual exclusivity association feature sequence in the target analysis window, and taking a ratio of the first frequency and the second frequency as an exclusive response intensity value; when the exclusive response intensity value is greater than the preset physical dominant threshold value, it is determined that the medium attenuation feature sequence has a sustained unidirectional change feature in the target analysis window, and a transmission medium physical abnormality tracing result is generated; when the exclusive response intensity value is less than the preset logical dominant threshold value, it is determined that the channel load competition state set has a stepwise mutation feature in the target analysis window, and a logical channel resource contention tracing result is generated; when the exclusive response intensity value is between the physical dominant threshold value and the logical dominant threshold value, if the medium attenuation feature sequence meets a unidirectional growth condition in the target analysis window, a transmission medium physical abnormality tracing result is generated, and if the channel load competition state set meets a stepwise mutation condition in the target analysis window, a logical channel resource contention tracing result is generated.

8. A microservice-oriented inter-container abnormal information tracing system, characterized in that, including: a generation module configured to generate a service communication performance data set based on a transmission rate indicator and an error rate indicator of an inter-container communication request collected; a collection module configured to collect an optical signal intensity physical quantity of a data transmission link carrying the inter-container communication request, and generate a medium attenuation feature sequence based on the optical signal intensity physical quantity; an analysis module configured to analyze a multiplexing logical structure of a shared channel between service instances to generate a channel load competition state set; a construction module configured to time-align the medium attenuation feature sequence and the channel load competition state set to construct a mutual exclusivity association feature sequence of a transmission medium physical state and a channel logical state; The analysis module is configured to, when abnormal fluctuation data appears in the service communication performance data set, calculate an exclusive response strength value according to the mutual exclusivity correlation feature sequence, compare the exclusive response strength value with a preset physical dominant threshold and a preset logical dominant threshold, and generate an abnormal information tracing result, wherein the abnormal information tracing result includes a transmission medium physical abnormality or a logical channel resource contention.

Images