Inter-process communication control method based on historical prior constraints

Abstract

This invention relates to the field of communication technology and provides an inter-process communication control method based on historical prior constraints. This method can profile behavior based on historical and current inter-process communication messages, calculate the risk prediction value for each message type based on its behavioral feature vector, and achieve quantitative modeling of message execution risk. It employs a starvation avoidance mechanism to adaptively adjust the scheduling priority of each message type based on its risk prediction value, and a load truncation mechanism to dynamically adjust the timeout threshold for each message type based on its behavioral feature vector, thus incorporating risk prediction results into scheduling and timeout control decisions. It scans the dead-letter record table at a preset period for cluster attribution analysis and uses a feedback mechanism to optimize the risk prediction function. This allows for reverse correction of the prediction model through structured analysis of dead-letter anomalies, integrating the scheduling mechanism and anomaly handling mechanism into a unified feedback control system.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to an inter-process communication control method based on historical prior constraints. Background Technology

[0002] In current operating systems, distributed computing platforms, and microservice architectures, inter-process communication (IPC) mechanisms undertake core functions such as task scheduling, message passing, and resource coordination. As system scale and business complexity increase, IPC systems typically exhibit the following characteristics during operation: (1) The message types are diverse, and the processing time for different types of tasks varies significantly; (2) The system load has obvious fluctuations, and the utilization rates of CPU (Central Processing Unit), memory and I / O (Input / Output) resources will change dynamically over time; (3) The message processing success rate is highly correlated with the system's operating status; (4) Queue backlog, timeout retries and abnormal propagation have a chain effect.

[0003] Furthermore, in existing technologies, IPC systems typically employ the following mechanisms: (1) Static scheduling strategy based on preset priority; (2) Fixed timeout threshold judgment mechanism; (3) Dead letter processing strategies based on frequency or rules; (4) The scheduling mechanism and the exception handling mechanism are independent of each other.

[0004] The above technical solutions can meet basic operational requirements when the system load is stable and the task characteristics are relatively simple. However, in environments with dynamic load fluctuations, the following shortcomings exist: (1) The scheduling decision lacks the use of historical behavior data and cannot reflect the risk differences of different message types; (2) The fixed timeout mechanism cannot adapt to the fluctuations in processing time, which may lead to misjudgment or delayed release of resources; (3) Dead letter analysis only focuses on the results and lacks structured modeling of the causes of anomalies; (4) There is no feedback mechanism between scheduling control and abnormal results, and the system cannot self-correct.

[0005] Under high load or abnormal propagation scenarios, the above deficiencies may lead to continuous queue backlog, intensified resource competition, increased number of timeouts, expansion of dead letters, and the system entering a vicious cycle of "congestion-timeout-retry-recongestion".

[0006] Therefore, existing IPC systems are essentially open-loop control structures, and their scheduling strategies cannot be dynamically adjusted based on historical operating behavior and abnormal results, which limits system stability and resource utilization efficiency. Summary of the Invention

[0007] In view of the above, it is necessary to provide an inter-process communication control method based on historical prior constraints, which aims to solve the problem that inter-process communication cannot be dynamically controlled based on historical priors.

[0008] An inter-process communication control method based on historical prior constraints, the method comprising: In response to inter-process communication control commands triggered at the current moment, retrieve historical inter-process communication messages and current inter-process communication messages; Based on the historical inter-process communication messages and the current inter-process communication messages, a behavioral profile is created to obtain a behavioral feature vector for each type of message. The constructed risk prediction function is invoked, and the risk prediction value for each type of message is calculated using the behavioral feature vector of each message type as the variable of the risk prediction function. A starvation avoidance mechanism is adopted to adaptively adjust the scheduling priority of each type of message based on the risk prediction value of each type of message, so as to obtain the current scheduling priority of each type of message. A load truncation mechanism is adopted to dynamically adjust the timeout threshold of each type of message based on the behavioral feature vector of each type of message, so as to obtain the current timeout threshold of each type of message. The subsequent inter-process communication messages are scheduled according to the current scheduling priority of each type of message, and dead-letter messages in the subsequent inter-process communication messages are detected according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message. Generate a dead-letter fingerprint for the dead-letter message and write the dead-letter fingerprint into the dead-letter record table; The dead letter record table is scanned according to a preset cycle to perform clustering attribution analysis and obtain the current dead letter type. A feedback mechanism is used to optimize the risk prediction function based on the current dead-letter type.

[0009] An inter-process communication control device based on historical prior constraints, the inter-process communication control device based on historical prior constraints includes: The acquisition unit is used to acquire historical inter-process communication messages and current inter-process communication messages in response to inter-process communication control instructions triggered at the current time. The profiling unit is used to perform behavioral profiling based on the historical inter-process communication messages and the current inter-process communication messages, and to obtain a behavioral feature vector for each type of message. The calculation unit is used to retrieve the constructed risk prediction function and calculate the risk prediction value for each type of message by using the behavioral feature vector of each message type as the variable of the risk prediction function. The adjustment unit is used to adaptively adjust the scheduling priority of each type of message based on the risk prediction value of each type of message using a starvation avoidance mechanism, so as to obtain the current scheduling priority of each type of message. The adjustment unit is also used to dynamically adjust the timeout threshold of each type of message based on the behavioral feature vector of each type of message using a load truncation mechanism, so as to obtain the current timeout threshold of each type of message. The detection unit is used to schedule subsequent inter-process communication messages according to the current scheduling priority of each type of message, and to detect dead-letter messages in the subsequent inter-process communication messages according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message. The generation unit is used to generate the dead letter fingerprint of the dead letter message and write the dead letter fingerprint into the dead letter record table; The analysis unit is used to scan the dead letter record table according to a preset period to perform cluster attribution analysis and obtain the current dead letter type; An optimization unit is used to optimize the risk prediction function based on the current dead-letter type using a feedback mechanism.

[0010] A computer device, the computer device comprising: A memory for storing at least one instruction; and a processor for executing the instructions stored in the memory to implement the inter-process communication control method based on historical prior constraints.

[0011] A computer-readable storage medium storing at least one instruction, which is executed by a processor in a computer device to implement the inter-process communication control method based on historical prior constraints.

[0012] As can be seen from the above technical solutions, this invention can perform behavioral profiling based on historical inter-process communication messages and current inter-process communication messages, and calculate the risk prediction value of each type of message using the behavioral feature vector of each type of message as the variable of the risk prediction function, thereby realizing quantitative modeling of message operation risk; it adopts a starvation avoidance mechanism to adaptively adjust the scheduling priority of each type of message according to the risk prediction value of each type of message, and adopts a load truncation mechanism to dynamically adjust the timeout threshold of each type of message according to the behavioral feature vector of each type of message, so as to introduce the risk prediction results into scheduling and timeout control decisions; it scans the dead letter record table according to a preset period for cluster attribution analysis, and uses a feedback mechanism to optimize the risk prediction function, and can reverse correct the prediction model through the structured analysis of dead letter anomalies, integrating the scheduling mechanism and the anomaly handling mechanism into a unified feedback control system. Attached Figure Description

[0013] Figure 1 This is a flowchart of a preferred embodiment of the inter-process communication control method based on historical prior constraints of the present invention; Figure 2 This is a functional block diagram of a preferred embodiment of the inter-process communication control device based on historical prior constraints of the present invention; Figure 3 This is a schematic diagram of the structure of a computer device that implements a preferred embodiment of the inter-process communication control method based on historical prior constraints according to the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0015] like Figure 1 The diagram shown is a flowchart of a preferred embodiment of the inter-process communication control method based on historical prior constraints according to the present invention. The order of the steps in this flowchart can be changed, and some steps can be omitted, depending on different requirements.

[0016] The inter-process communication control method based on historical prior constraints is applied to one or more computer devices. The computer device is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0017] The computer device can be any electronic product that can interact with the user, such as a personal computer, tablet computer, smartphone, personal digital assistant (PDA), interactive network television (IPTV), smart wearable device, etc.

[0018] The computer equipment may also include network equipment and / or user equipment. The network equipment includes, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of hosts or network servers.

[0019] The server can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0020] Artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0021] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0022] The network in which the computer device is located includes, but is not limited to, the Internet, wide area network, metropolitan area network, local area network, and virtual private network (VPN).

[0023] S10, in response to the inter-process communication control instruction triggered at the current moment, obtains historical inter-process communication messages and current inter-process communication messages.

[0024] In this embodiment, the inter-process communication control command can be triggered periodically.

[0025] In this embodiment, the current inter-process communication message is the inter-process communication message received at the current moment.

[0026] S11, Based on the historical inter-process communication messages and the current inter-process communication messages, a behavioral profile is created to obtain a behavioral feature vector for each type of message.

[0027] In this embodiment, the step of creating a behavioral profile based on the historical inter-process communication messages and the current inter-process communication messages to obtain a behavioral feature vector for each type of message includes: For each type of message, obtain the current message instance of that type from the current inter-process communication messages; Using a smoothing coefficient, the historical average processing time and processing time variance of this type of message are calculated based on the current message instance and the historical inter-process communication messages. Calculate the historical failure rate and historical timeout rate of this type of message based on the historical inter-process communication messages. The current time is used to determine the number of messages of this type in the waiting queue as the queue depth. The CPU (Central Processing Unit) utilization, CPU wait time for I / O (Input / Output) ratio, and memory utilization at the current moment are obtained, and the CPU utilization, CPU wait time for I / O ratio, and memory utilization are weighted and calculated to obtain the comprehensive system load index value. Based on a smoothing constant, the historical average processing time, the processing time variance, the historical failure rate, the historical timeout ratio, the queuing depth, and the system comprehensive load index value are normalized to obtain the behavioral feature vector of this type of message.

[0028] Inter-process communication messages may include, but are not limited to: (1) Control signaling messages: used for inter-process state synchronization, instruction issuance and health detection, such as heartbeat packets, configuration synchronization instructions, fault switching signaling, etc. These messages have high priority and are sensitive to delay; (2) Business data messages: These are used to carry core business interaction data, such as business requests, response messages, batch data transmission messages, etc. These messages have high requirements for processing time and throughput. (3) Event notification messages: used for asynchronous event triggering and pushing, such as alarm event notifications, status change notifications, etc. These messages have certain requirements for timely delivery; (4) System management messages: used for background management scenarios such as process registration, resource scheduling and permission verification; (5) Debugging and monitoring messages: used for non-core business scenarios such as log collection and performance probes. These messages have a high tolerance for latency.

[0029] Specifically, the historical average processing time represents the historical average execution time for each type of message, and can be calculated using the following formula: ; in, λ represents the historical average processing time of the i-th type of message calculated at the current time t; λ represents the smoothing coefficient, with a value range of 0 < λ ≤ 1. The larger λ is, the more sensitive it is to new data response. The default value of λ can be 0.2. This indicates the actual processing time of the current message instance from the start of its execution to its completion (which can be obtained by calculating the difference between the system timestamp when the message is completed and returns a result and the system timestamp when the message enters the execution thread pool; the time unit can be milliseconds, and the system timestamp can be derived from the operating system's high-precision timer). This represents the historical average processing time of message type i calculated at time t-1.

[0030] Specifically, the processing time variance is used to reflect the stability of the processing time for this message type, and can be calculated using the following formula: ; in, This represents the variance of processing time for the i-th type of message at the current time t, and is initialized at... ; Let represent the variance of processing time for message type i at time t-1.

[0031] Specifically, the historical failure rate is used to represent the failure rate of a certain message type, and can be calculated using the following formula: ; in, This represents the historical failure rate of the i-th type of message; This represents the number of failures for the i-th type of message (e.g., the number of times an error code is returned, an exception is thrown, or a timeout occurs). This represents the cumulative number of times the i-th type of message has been executed.

[0032] Furthermore, to prevent excessive fluctuations in initial data, the following protection mechanisms can be adopted: when hour, Use the default value of 0.1; when hour, ; in, This represents the minimum sample size threshold, which can be configured to be 30.

[0033] Specifically, the historical timeout ratio can be calculated using the following formula: ; in, This represents the historical timeout percentage for the i-th type of message; This represents the number of timeouts for the i-th type of message, i.e., the number of times the execution time exceeds the current timeout threshold.

[0034] Specifically, the queuing depth can be calculated using the following formula: ; in, This represents the queuing depth of the i-th type of message; This represents the waiting queue for message type i. The length of the queue, i.e. the number of elements in the queue.

[0035] Specifically, the system comprehensive load index value is a normalized index of the system load, which can be calculated using the following formula: ; in, This represents the overall load index value of the system; This represents the normalized CPU utilization, and , This indicates the percentage of CPU utilization. This represents the proportion of CPU time spent waiting for I / O after normalization, and , This indicates the proportion of time the CPU spends waiting for I / O. This represents the normalized memory utilization rate. , Indicates the amount of memory already used. Indicates the total memory capacity; , , Represents the weighting coefficient, and ,like , , .

[0036] These parameters together constitute a statistical feature vector of message execution risk, which can be used to transform discrete execution behavior into a computable statistical structure.

[0037] Furthermore, the historical average processing time, the processing time variance, the historical failure rate, the historical timeout ratio, the queuing depth, and the overall system load index can be normalized using the following normalization functions: ; Where x represents any one of the historical average processing time, the processing time variance, the historical failure rate, the historical timeout ratio, the queuing depth, and the system comprehensive load index value.

[0038] k represents the smoothing constant, used to avoid the denominator approaching 0 due to small samples. The rules for choosing the value of k are as follows: If x is a proportional variable, then k = 0.1; If x is a variance variable, then k = 10% of the historical mean; If x is the queuing depth, then k = 10.

[0039] In the general scenario, k=1.

[0040] Through the above data modeling, we can abstract message behavior, no longer just look at the current message, but build a behavioral profile for each type of message based on historical priors.

[0041] In this embodiment, when the system is initialized, the historical average processing time can be assigned an initial value of 200ms, and the processing time variance can be assigned an initial value of 50ms. 2 The initial value of the historical failure rate is 0.1, and the initial value of the historical timeout ratio is 0.05.

[0042] In this embodiment, relevant indicators can be screened for relevance (e.g., calculating the Pearson correlation coefficient between each indicator and the target variable) and ranked for importance (e.g., using random forest to calculate the feature importance of each indicator) based on historical data, so as to obtain indicators (e.g., the historical timeout ratio, the queuing depth, etc.) to construct the behavioral feature vector of each type of message.

[0043] S12, retrieve the constructed risk prediction function, and calculate the risk prediction value for each type of message by using the behavioral feature vector of each message type as the variable of the risk prediction function.

[0044] In this embodiment, the risk prediction function can be expressed as follows: ; in, This represents the predicted risk value for the i-th type of message; α, β, γ, δ, and ε are adjustable weight parameters that can be automatically optimized based on historical replay data, satisfying: α + β + γ + δ + ε = 1. Initially, α = 0.25, representing the failure rate weight; β = 0.2, representing the variance weight; γ = 0.2, representing the queuing depth weight; δ = 0.2, representing the system load weight; and ε = 0.15, representing the timeout ratio weight.

[0045] Furthermore, the calculation of the risk prediction value for each type of message by using the behavioral feature vector of each message type as the variable in the risk prediction function includes: Obtain the configured weight coefficient matrix; The weighted sum of the behavioral feature vectors of each type of message is calculated based on the weight coefficient matrix to obtain the risk prediction value of each type of message.

[0046] The risk prediction value can be used to characterize the probability trend of the current message becoming abnormal in the current environment.

[0047] Through the above embodiments, the behavioral feature vector of each type of message can be transformed into a unified decision-making basis, namely the risk prediction value. The risk prediction value can serve as a unique decision signal for the entire system, driving subsequent scheduling control, timeout control, and dead-letter attribution, enabling message scheduling and exception handling to be based on the same prediction basis for the first time.

[0048] S13 employs a starvation avoidance mechanism, adaptively adjusting the scheduling priority of each type of message based on the risk prediction value of each message type to obtain the current scheduling priority of each message type.

[0049] In this embodiment, the starvation avoidance mechanism is adopted to adaptively adjust the scheduling priority of each type of message based on the risk prediction value of each type of message, and the current scheduling priority of each type of message is obtained by: For each type of message, obtain the baseline scheduling priority, risk amplification factor, and aging decay factor for that type of message based on the business attribute configuration; Calculate the risk amplification factor and the risk prediction value of this type of message to obtain the risk compensation value of this type of message; Obtain the average waiting time for this type of message to be invoked in the waiting queue; An aging factor is generated based on the average waiting time. Calculate the product of the aging decay coefficient and the aging factor to obtain the starvation avoidance adjustment value for this type of message; The current scheduling priority of this type of message is obtained by summing the baseline scheduling priority, the risk compensation value, and the starvation avoidance adjustment value.

[0050] Specifically, the aging factor can be calculated using the following formula: ; in, This refers to the aging factor; This indicates the average waiting time.

[0051] The baseline scheduling priority can range from 1 to 10.

[0052] The risk amplification factor can range from 0 to 5, with a default value of 2.

[0053] The aging degradation coefficient can be 0.5.

[0054] All of the above values ​​can be adjusted according to the actual needs of the scenario, and the final calculation result can be guaranteed to be in integer priority form by rounding.

[0055] In the above embodiments, by introducing an aging factor, the long-term starvation of low-priority messages can be avoided, and the priority of messages with longer waiting times can be gradually increased.

[0056] S14. A load truncation mechanism is adopted to dynamically adjust the timeout threshold of each type of message based on the behavioral feature vector of each type of message, so as to obtain the current timeout threshold of each type of message.

[0057] In this embodiment, the load truncation mechanism dynamically adjusts the timeout threshold for each message type based on its behavioral feature vector, resulting in the current timeout threshold for each message type, including: For each type of message, obtain the confidence coefficient and load factor of that type of message based on the business attribute configuration; Calculate the standard deviation of processing time for this type of message based on its behavioral feature vector; The fluctuation compensation value is obtained by multiplying the confidence coefficient by the standard deviation of the processing time. Calculate the sum of the historical average processing time of this type of message and the fluctuation compensation value to obtain the basic timeout threshold of this type of message; The basic timeout threshold is corrected based on the load factor to obtain the current timeout threshold for this type of message; The load factor is the smaller value between the system comprehensive load index value of this type of message and 1, and the maximum value of the load factor is truncated to 1, so as to ensure that the timeout threshold will not expand indefinitely, and that the dynamic timeout threshold remains within a reasonable and controllable range under high load scenarios, avoiding the long-term occupation of resources due to excessively large thresholds, thereby maintaining the effectiveness and stability of the system resource reclamation mechanism.

[0058] Specifically, the current timeout threshold can be calculated using the following formula: ; in, This represents the current timeout threshold for the i-th type of message; represents the standard deviation of the processing time of the i-th type of message at the current time t; k1 represents the confidence coefficient, which can be 2, corresponding to an approximately 95% confidence interval.

[0059] S15, schedule subsequent inter-process communication messages according to the current scheduling priority of each type of message, and detect dead-letter messages in the subsequent inter-process communication messages according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message.

[0060] In this embodiment, after adjusting the priority based on the risk prediction results, subsequent message scheduling can be performed according to the adjusted priority.

[0061] Similarly, after adjusting the timeout threshold based on the risk prediction results, dead-letter messages can also be detected based on the adjusted timeout threshold.

[0062] In this embodiment, detecting dead-letter messages in subsequent inter-process communication messages based on the current timeout threshold of each message type throughout the entire lifecycle of each message type includes: When the execution time of an inter-process communication message is detected to be greater than the corresponding current timeout threshold, the detected inter-process communication message is determined to be the dead letter message.

[0063] S16, generate the dead letter fingerprint of the dead letter message, and write the dead letter fingerprint into the dead letter record table.

[0064] In this embodiment, generating the dead-letter fingerprint of the dead-letter message includes: Obtain the message type identifier, failure stage identifier, system resource load index value, network round-trip delay, and risk prediction value of the dead letter message to construct an original continuous feature vector; The original continuous feature vector is discretized into discrete intervals to obtain discrete interval encoded features; Calculate the hash value of the discrete interval encoded feature as a fingerprint index; The dead-letter fingerprint is obtained by combining the original continuous feature vector, the discrete interval encoded features, and the fingerprint index.

[0065] The system resource load metrics may include average CPU utilization, average memory usage, and I / O wait rate during processing.

[0066] The original continuous feature vector is a continuous numerical space. In order to improve the clustering stability and dimensionality reduction efficiency, the continuous features need to be discretized into intervals.

[0067] For example, the discrete interval coding feature can be represented as: ; in, The discrete interval encoding feature is represented by Type; the message type identifier is represented by Stage; and the processing stage identifier for the failure is represented by Stage. This indicates the interval number to which the average CPU utilization during the processing period belongs; This indicates the interval number to which the average memory usage rate belongs; This indicates the interval number to which the I / O wait ratio belongs; Indicates the interval number to which the network round-trip delay belongs; This indicates the interval number to which the predicted risk value belongs.

[0068] The discrete interval coding feature belongs to the discrete space feature.

[0069] Specifically, when mapping the corresponding original continuous feature vectors to interval numbers through interval discretization, the following method can be used: (1) For the average CPU utilization during the processing period, a fixed interval segmentation method can be used: For example: P1: 0% ≤ average CPU utilization during processing < 40%; P2: 40% ≤ Average CPU utilization during processing < 70%; P3: 70% ≤ Average CPU utilization during processing < 85%; P4: 85% ≤ Average CPU utilization during processing ≤ 100%.

[0070] The threshold values ​​for each interval can be configured based on the number of system cores and the load model.

[0071] (2) For the average memory usage rate, a relative usage rate segmentation method can be used: For example: M1: 0% ≤ average memory usage < 50%; M2: 50% ≤ Average memory utilization < 75%; M3: 75% ≤ Average memory utilization < 90%; M4: 90% ≤ average memory usage ≤ 100%.

[0072] Among them, when the average memory utilization rate is ≥90%, it is considered a high memory pressure state.

[0073] (3) The I / O wait ratio can be segmented as follows: For example: I1: 0%≤I / O wait rate<10%; I2: 10% ≤ I / O wait rate < 25%; I3: 25% ≤ I / O wait rate < 50%; I4: 50% ≤ I / O wait rate ≤ 100%.

[0074] When the I / O wait rate is ≥50%, it is judged as a severe I / O blocking state.

[0075] (4) For the network round-trip delay, an adaptive baseline segmentation method can be used: Calculate the historical mean and standard deviation based on historical network round-trip delays, and then segment the network based on the historical mean and standard deviation: For example: N1: Network round-trip latency < historical average; N2: Historical mean ≤ network round-trip delay < historical mean + historical standard deviation; N3: Historical mean + historical standard deviation ≤ network round-trip delay < historical mean + 2 * historical standard deviation; N4: Network round-trip latency ≥ historical mean + 2 * historical standard deviation.

[0076] The above segmentation method is applicable to different network environments.

[0077] (5) For the predicted risk value, the following segmentation method can be used: For example: R1: 0 ≤ predicted risk value < 0.25; R²: 0.25 ≤ predicted risk value < 0.5; R3: 0.5 ≤ predicted risk value < 0.75; R4: 0.75 ≤ predicted risk value ≤ 1.

[0078] Furthermore, to facilitate rapid retrieval, clustering, and analysis, the discrete interval coding features can be uniformly encoded to generate the fingerprint index.

[0079] For example, the message type identifier, the failure processing stage identifier, the interval number of the average CPU utilization during processing, the interval number of the average memory usage, the interval number of the I / O wait ratio, the interval number of the network round-trip delay, and the interval number of the risk prediction value can be concatenated into strings in sequence, and the hash value of the concatenated string can be calculated using hash algorithms such as Murmur Hash, SHA-1 (Secure Hash Algorithm 1), and CRC32 (Cyclic Redundancy Check 32) to serve as the fingerprint index.

[0080] The output length of the fingerprint index can be configured according to actual needs. Furthermore, the hash value is only used for indexing and storage optimization and does not change the clustering basis of the features themselves.

[0081] In the above embodiments, the dead-letter fingerprint is generated based on multi-dimensional structured coding, which can accurately distinguish between structural anomalies and transient anomalies, and realize automatic anomaly location.

[0082] In this embodiment, graded blocking can also be performed based on the number of consecutive failures of the same dead-letter fingerprint.

[0083] Specifically, when messages corresponding to the same dead-letter fingerprint fail consecutively ≥ a first preset number of times (e.g., 3 times), the priority of such messages can be reduced and the retry interval delayed to limit resource consumption while preserving business attempt opportunities.

[0084] When the message corresponding to the same dead-letter fingerprint fails for more than or equal to the second preset number of times (e.g., 5 times), the scheduling of such messages can be paused, and only low-frequency liveness detection (e.g., 1 time / minute) can be retained to greatly reduce invalid retries and break the vicious cycle.

[0085] When a message corresponding to the same dead-letter fingerprint fails for a third preset number of times (e.g., 10 times), the message can be rejected from entering the waiting queue and written directly to the dead-letter queue to completely block the propagation of anomalies and protect the core business of the system.

[0086] S17. Scan the dead letter record table according to a preset cycle to perform cluster attribution analysis and obtain the current dead letter type.

[0087] In this embodiment, the step of scanning the dead letter record table according to a preset period to perform cluster attribution analysis and obtain the current dead letter type includes: Obtain a first threshold and a second threshold; wherein the first threshold is greater than the second threshold; Obtain the total number of dead letters scanned, and the number of fingerprints for each dead letter; The frequency of occurrence of each dead-letter fingerprint is obtained by calculating the quotient of the number of each dead-letter fingerprint to the total number of dead letters; For each dead-letter fingerprint, when the frequency of occurrence of the dead-letter fingerprint is greater than the first threshold, the corresponding current dead-letter type is determined to be a structural anomaly type; or When the frequency of occurrence of the dead letter fingerprint is less than the second threshold, the corresponding current dead letter type is determined to be a transient anomaly type.

[0088] The preset period, the first threshold, and the second threshold can be comprehensively configured and adjusted based on business needs, system performance, and historical data. For example, the first threshold can be configured to a value greater than 0.3, and the second threshold can be configured to a value less than 0.05.

[0089] In this embodiment, different response measures can be taken for different current dead letter types.

[0090] For example, when the current dead-letter type is the structural anomaly type, it indicates that the anomaly is caused by fixed factors such as inherent message defects, system resource bottlenecks, and architectural configuration errors, exhibiting characteristics of concentrated outbreaks and continuous recurrence. In this case, closed-loop feedback correction can be triggered, increasing the failure rate weight in the risk model, improving the risk score of similar messages, and reducing the scheduling priority of this type of message to reduce its system resource consumption. Simultaneously, rate limiting or circuit breaker control can be implemented for this type of message to block anomaly propagation and prevent the system from entering a congestion avalanche. Furthermore, reporting anomaly information facilitates the location and elimination of underlying fault root causes.

[0091] When the current dead letter type is the transient anomaly type, it indicates that the anomaly is caused by occasional factors such as instantaneous system fluctuations, network jitter, and temporary resource preemption, and does not have regularity or repetition. In this case, the message can be retried according to the conventional retry strategy, while only logging is performed and rate limiting or circuit breaking mechanisms are not activated to ensure the normal operation of the business.

[0092] In this embodiment, after obtaining the occurrence frequency of each dead-letter fingerprint, the method further includes: For any dead-letter fingerprint, obtain all target dead-letter messages corresponding to the arbitrary dead-letter fingerprint; When the system comprehensive load index value corresponding to more than a preset proportion of dead letters in the target dead letters is greater than a preset threshold, a resource bottleneck abnormal event is determined to have occurred. Increase the load factor of the message type corresponding to the arbitrary dead-letter fingerprint to relax the corresponding current timeout threshold and reduce the corresponding current scheduling priority.

[0093] The preset ratio and the preset threshold can also be configured and adjusted comprehensively according to business needs, system performance and historical data.

[0094] Through the above embodiments, when high-frequency fingerprints are concentrated in the high-load range, it can be determined that there is an abnormality in the resource bottleneck, and an effective response to high-load scenarios can be achieved.

[0095] S18, a feedback mechanism is adopted to optimize the risk prediction function based on the current dead letter type.

[0096] In this embodiment, the step of using a feedback mechanism to optimize the risk prediction function based on the current dead-letter type includes: Determine the feedback score based on the current dead letter type; Obtain the learning rate and calculate the product of the learning rate and the score to obtain the adjustment step size; The weight coefficient matrix in the risk prediction function is adjusted according to the adjustment step size.

[0097] For example, for the weight coefficient α in the weight coefficient matrix, the sum of α and the adjustment step size can be calculated to obtain a new weight coefficient α.

[0098] When the current dead letter type is a structural anomaly type, the feedback score can be 1 (which can slowly increase the α equal weight coefficient, increase the risk prediction value of similar messages, reduce the priority, and make them more likely to be limited or circuit-breakered). When the current dead-letter type is a transient anomaly, the feedback score can be -1, -0.1, or 0 (this allows the α-weighted coefficient to decrease slowly, avoiding oversensitivity to occasional failures and preventing the false positive of normal messages). In the case of transient anomalies, the feedback score is generally set to a relatively small value, with only slight adjustments, to prevent the risk prediction model (i.e., the risk prediction function) from becoming overly sensitive. Simultaneously, it avoids the situation where, if a large number of transient anomalies occur, the weighted coefficients are continuously suppressed, causing the model to become insensitive to subsequent real structural anomalies, resulting in passivation.

[0099] Among these features, maximum and minimum value constraints can be configured for each weight coefficient to prevent it from continuously increasing or decreasing, ensuring that the weight coefficient remains within a reasonable range.

[0100] The learning rate can be 0.01.

[0101] Of course, the behavioral feature vector, risk prediction value, and priority of each type of message can be further updated.

[0102] Through the above embodiments, the weight coefficients can be adjusted according to the market type to achieve continuous optimization.

[0103] In this embodiment, when a dead letter message occurs, it not only records the error but also clusters the dead letters according to the operating environment and risk level to determine whether it is a structural anomaly or an occasional anomaly. The anomaly results are then fed back to the risk model to adjust the weights and parameters, and the anomaly results are transformed into model correction signals. This achieves a closed-loop mode from statistical model to scheduling control, to anomaly results, and then to model correction (the functional modules corresponding to each link can be deployed in the IPC scheduling service process). This reduces the probability of misjudging dead letters, reduces system blockage, and improves the accuracy of anomaly attribution, thereby improving the operational stability and resource utilization efficiency of the IPC system under dynamic load environments.

[0104] As can be seen from the above technical solutions, this invention can perform behavioral profiling based on historical inter-process communication messages and current inter-process communication messages, and calculate the risk prediction value of each type of message using the behavioral feature vector of each type of message as the variable of the risk prediction function, thereby realizing quantitative modeling of message operation risk; it adopts a starvation avoidance mechanism to adaptively adjust the scheduling priority of each type of message according to the risk prediction value of each type of message, and adopts a load truncation mechanism to dynamically adjust the timeout threshold of each type of message according to the behavioral feature vector of each type of message, so as to introduce the risk prediction results into scheduling and timeout control decisions; it scans the dead letter record table according to a preset period for cluster attribution analysis, and uses a feedback mechanism to optimize the risk prediction function, and can reverse correct the prediction model through the structured analysis of dead letter anomalies, integrating the scheduling mechanism and the anomaly handling mechanism into a unified feedback control system.

[0105] like Figure 2The diagram shown is a functional block diagram of a preferred embodiment of the inter-process communication control device based on historical prior constraints of the present invention. The inter-process communication control device 11 based on historical prior constraints includes an acquisition unit 110, a profiling unit 111, a calculation unit 112, an adjustment unit 113, a detection unit 114, a generation unit 115, an analysis unit 116, and an optimization unit 117. The module / unit referred to in this invention refers to a series of computer program segments that can be executed by a processor and perform a fixed function, and are stored in memory. In this embodiment, the functions of each module / unit will be described in detail in subsequent embodiments.

[0106] The acquisition unit 110 is used to acquire historical inter-process communication messages and current inter-process communication messages in response to an inter-process communication control command triggered at the current time. The profiling unit 111 is used to perform behavioral profiling based on the historical inter-process communication messages and the current inter-process communication messages to obtain a behavioral feature vector for each type of message. The calculation unit 112 is used to retrieve the constructed risk prediction function and calculate the risk prediction value of each type of message by using the behavioral feature vector of each type of message as the variable of the risk prediction function. The adjustment unit 113 is used to adopt a starvation avoidance mechanism to adaptively adjust the scheduling priority of each type of message according to the risk prediction value of each type of message, so as to obtain the current scheduling priority of each type of message. The adjustment unit 113 is also used to dynamically adjust the timeout threshold of each type of message according to the behavioral feature vector of each type of message using a load truncation mechanism, so as to obtain the current timeout threshold of each type of message. The detection unit 114 is used to schedule subsequent inter-process communication messages according to the current scheduling priority of each type of message, and to detect dead-letter messages in the subsequent inter-process communication messages according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message. The generation unit 115 is used to generate the dead letter fingerprint of the dead letter message and write the dead letter fingerprint into the dead letter record table; The analysis unit 116 is used to scan the dead letter record table according to a preset period to perform cluster attribution analysis and obtain the current dead letter type. The optimization unit 117 is used to optimize the risk prediction function based on the current dead-letter type using a feedback mechanism.

[0107] As can be seen from the above technical solutions, this invention can perform behavioral profiling based on historical inter-process communication messages and current inter-process communication messages, and calculate the risk prediction value of each type of message using the behavioral feature vector of each type of message as the variable of the risk prediction function, thereby realizing quantitative modeling of message operation risk; it adopts a starvation avoidance mechanism to adaptively adjust the scheduling priority of each type of message according to the risk prediction value of each type of message, and adopts a load truncation mechanism to dynamically adjust the timeout threshold of each type of message according to the behavioral feature vector of each type of message, so as to introduce the risk prediction results into scheduling and timeout control decisions; it scans the dead letter record table according to a preset period for cluster attribution analysis, and uses a feedback mechanism to optimize the risk prediction function, and can reverse correct the prediction model through the structured analysis of dead letter anomalies, integrating the scheduling mechanism and the anomaly handling mechanism into a unified feedback control system.

[0108] like Figure 3 The diagram shown is a structural schematic of a computer device that implements a preferred embodiment of the inter-process communication control method based on historical prior constraints according to the present invention.

[0109] The computer device 1 may include a memory 12, a processor 13, and a bus (the arrow in the figure represents the bus), and may also include a computer program stored in the memory 12 and executable on the processor 13, such as an inter-process communication control program based on historical prior constraints.

[0110] Those skilled in the art will understand that the schematic diagram is merely an example of computer device 1 and does not constitute a limitation on computer device 1. Computer device 1 can be either a bus topology or a star topology. Computer device 1 may also include more or fewer other hardware or software than shown in the diagram, or different component arrangements. For example, computer device 1 may also include input / output devices, network access devices, etc.

[0111] It should be noted that the computer device 1 described is merely an example. Other existing or future electronic products that are adaptable to this invention should also be included within the scope of protection of this invention and are incorporated herein by reference.

[0112] The memory 12 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 12 can be an internal storage unit of the computer device 1, such as a portable hard drive of the computer device 1. In other embodiments, the memory 12 can be an external storage device of the computer device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device 1. Furthermore, the memory 12 can include both internal and external storage units of the computer device 1. The memory 12 can be used not only to store application software and various types of data installed on the computer device 1, such as the code of inter-process communication control programs based on historical prior constraints, but also to temporarily store data that has been output or will be output.

[0113] In some embodiments, the processor 13 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits packaged with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 13 is the control unit of the computer device 1, connecting various components of the computer device 1 via various interfaces and lines. It executes programs or modules stored in the memory 12 (e.g., executing inter-process communication control programs based on historical prior constraints) and calls data stored in the memory 12 to perform various functions of the computer device 1 and process data.

[0114] The processor 13 executes the operating system of the computer device 1 and various installed applications. The processor 13 executes the applications to implement the steps in the above embodiments of the inter-process communication control method based on historical prior constraints, for example... Figure 1 The steps are shown.

[0115] For example, the computer program can be divided into one or more modules / units, which are stored in the memory 12 and executed by the processor 13 to complete the present invention. The one or more modules / units can be a series of computer-readable instruction segments capable of performing specific functions, which describe the execution process of the computer program in the computer device 1. For example, the computer program can be divided into an acquisition unit 110, an image processing unit 111, a calculation unit 112, an adjustment unit 113, a detection unit 114, a generation unit 115, an analysis unit 116, and an optimization unit 117.

[0116] The integrated unit implemented as a software functional module described above can be stored in a computer-readable storage medium. This software functional module, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, a computer device, or a network device, etc.) or processor to execute the inter-process communication control method based on historical prior constraints described in various embodiments of the present invention.

[0117] If the modules / units integrated in the computer device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware devices. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above.

[0118] The computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory, etc.

[0119] Furthermore, the computer-readable storage medium may primarily include a stored program area and a stored data area, wherein the stored program area may store the operating system, an application program required for at least one function, etc.; and the stored data area may store data created based on the use of blockchain nodes, etc.

[0120] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0121] The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, in... Figure 3 The bus is represented by only one straight line, but this does not mean that there is only one bus or one type of bus. The bus is configured to enable communication between the memory 12 and at least one processor 13, etc.

[0122] Although not shown, the computer device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 13 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The computer device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0123] Furthermore, the computer device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish a communication connection between the computer device 1 and other computer devices.

[0124] Optionally, the computer device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the computer device 1 and to display a visual user interface.

[0125] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0126] It will be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the computer device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0127] Combination Figure 1 The memory 12 in the computer device 1 stores multiple instructions to implement an inter-process communication control method based on historical prior constraints, and the processor 13 can execute the multiple instructions to achieve: In response to inter-process communication control commands triggered at the current moment, retrieve historical inter-process communication messages and current inter-process communication messages; Based on the historical inter-process communication messages and the current inter-process communication messages, a behavioral profile is created to obtain a behavioral feature vector for each type of message. The constructed risk prediction function is invoked, and the risk prediction value for each type of message is calculated using the behavioral feature vector of each message type as the variable of the risk prediction function. A starvation avoidance mechanism is adopted to adaptively adjust the scheduling priority of each type of message based on the risk prediction value of each type of message, so as to obtain the current scheduling priority of each type of message. A load truncation mechanism is adopted to dynamically adjust the timeout threshold of each type of message based on the behavioral feature vector of each type of message, so as to obtain the current timeout threshold of each type of message. The subsequent inter-process communication messages are scheduled according to the current scheduling priority of each type of message, and dead-letter messages in the subsequent inter-process communication messages are detected according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message. Generate a dead-letter fingerprint for the dead-letter message and write the dead-letter fingerprint into the dead-letter record table; The dead letter record table is scanned according to a preset cycle to perform clustering attribution analysis and obtain the current dead letter type. A feedback mechanism is used to optimize the risk prediction function based on the current dead-letter type.

[0128] Specifically, the processor 13's implementation method for the above instructions can be found in [reference needed]. Figure 1 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0129] It should be noted that all the data involved in this case was legally obtained.

[0130] If any AI models, software tools, or components not belonging to this company appear in the embodiments of this invention, they are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this invention has been obtained by an entity authorized (with the knowledge and consent) or fully authorized by all parties through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0131] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0132] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0133] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0134] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0135] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0136] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0137] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices described in this invention can also be implemented by a single unit or device through software or hardware. Terms such as "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for inter-process communication control based on historical prior constraints, characterized in that, The inter-process communication control method based on historical prior constraints includes: In response to inter-process communication control commands triggered at the current moment, retrieve historical inter-process communication messages and current inter-process communication messages; Based on the historical inter-process communication messages and the current inter-process communication messages, a behavioral profile is created to obtain a behavioral feature vector for each type of message. The constructed risk prediction function is invoked, and the risk prediction value for each type of message is calculated using the behavioral feature vector of each message type as the variable of the risk prediction function. A starvation avoidance mechanism is adopted to adaptively adjust the scheduling priority of each type of message based on the risk prediction value of each type of message, so as to obtain the current scheduling priority of each type of message. A load truncation mechanism is adopted to dynamically adjust the timeout threshold of each type of message based on the behavioral feature vector of each type of message, so as to obtain the current timeout threshold of each type of message. The subsequent inter-process communication messages are scheduled according to the current scheduling priority of each type of message, and dead-letter messages in the subsequent inter-process communication messages are detected according to the current timeout threshold of each type of message throughout the entire life cycle of each type of message. Generate a dead-letter fingerprint for the dead-letter message and write the dead-letter fingerprint into the dead-letter record table; The dead letter record table is scanned according to a preset cycle to perform clustering attribution analysis and obtain the current dead letter type. A feedback mechanism is used to optimize the risk prediction function based on the current dead-letter type.

2. The inter-process communication control method based on historical prior constraints as described in claim 1, characterized in that, The behavioral profile obtained by performing behavioral profiling based on the historical inter-process communication messages and the current inter-process communication messages includes: For each type of message, obtain the current message instance of that type from the current inter-process communication messages; Using a smoothing coefficient, the historical average processing time and processing time variance of this type of message are calculated based on the current message instance and the historical inter-process communication messages. Calculate the historical failure rate and historical timeout rate of this type of message based on the historical inter-process communication messages. The current time is used to determine the number of messages of this type in the waiting queue as the queue depth. The CPU utilization, CPU wait time for I / O, and memory utilization at the current moment are obtained, and the CPU utilization, CPU wait time for I / O, and memory utilization are weighted and calculated to obtain the comprehensive system load index value. Based on a smoothing constant, the historical average processing time, the processing time variance, the historical failure rate, the historical timeout ratio, the queuing depth, and the system comprehensive load index value are normalized to obtain the behavioral feature vector of this type of message.

3. The inter-process communication control method based on historical prior constraints as described in claim 1, characterized in that, The calculation, which uses the behavioral feature vector of each message type as the variable in the risk prediction function, yields the risk prediction value for each message type, including: Obtain the configured weight coefficient matrix; The weighted sum of the behavioral feature vectors of each type of message is calculated based on the weight coefficient matrix to obtain the risk prediction value of each type of message.

4. The inter-process communication control method based on historical prior constraints as described in claim 2, characterized in that, The starvation avoidance mechanism adaptively adjusts the scheduling priority of each message type based on the risk prediction value of each message type, resulting in the current scheduling priority of each message type, including: For each type of message, obtain the baseline scheduling priority, risk amplification factor, and aging decay factor for that type of message based on the business attribute configuration; Calculate the risk amplification factor and the risk prediction value of this type of message to obtain the risk compensation value of this type of message; Obtain the average waiting time for this type of message to be invoked in the waiting queue; An aging factor is generated based on the average waiting time. Calculate the product of the aging decay coefficient and the aging factor to obtain the starvation avoidance adjustment value for this type of message; The current scheduling priority of this type of message is obtained by summing the baseline scheduling priority, the risk compensation value, and the starvation avoidance adjustment value.

5. The inter-process communication control method based on historical prior constraints as described in claim 2, characterized in that, The load truncation mechanism dynamically adjusts the timeout threshold for each message type based on its behavioral feature vector. The current timeout threshold for each message type includes: For each type of message, obtain the confidence coefficient and load factor of that type of message based on the business attribute configuration; Calculate the standard deviation of processing time for this type of message based on its behavioral feature vector; The fluctuation compensation value is obtained by multiplying the confidence coefficient by the standard deviation of the processing time. Calculate the sum of the historical average processing time of this type of message and the fluctuation compensation value to obtain the basic timeout threshold of this type of message; The basic timeout threshold is corrected based on the load factor to obtain the current timeout threshold for this type of message; The load factor is the smaller value between the system comprehensive load index value of this type of message and 1, and the maximum value of the load factor is truncated to 1.

6. The inter-process communication control method based on historical prior constraints as described in claim 1, characterized in that, The step of detecting dead-letter messages in subsequent inter-process communication messages based on the current timeout threshold of each message type throughout the entire lifecycle includes: When the execution time of an inter-process communication message is detected to be greater than the corresponding current timeout threshold, the detected inter-process communication message is determined to be the dead letter message.

7. The inter-process communication control method based on historical prior constraints as described in claim 1, characterized in that, The dead-letter fingerprint used to generate the dead-letter message includes: Obtain the message type identifier, failure stage identifier, system resource load index value, network round-trip delay, and risk prediction value of the dead letter message to construct an original continuous feature vector; The original continuous feature vector is discretized into discrete intervals to obtain discrete interval encoded features; Calculate the hash value of the discrete interval encoded feature as a fingerprint index; The dead-letter fingerprint is obtained by combining the original continuous feature vector, the discrete interval encoded features, and the fingerprint index.

8. The inter-process communication control method based on historical prior constraints as described in claim 5, characterized in that, The step of scanning the dead letter record table according to a preset period to perform cluster attribution analysis yields the following current dead letter types: Obtain a first threshold and a second threshold; wherein the first threshold is greater than the second threshold; Obtain the total number of dead letters scanned, and the number of fingerprints for each dead letter; The frequency of occurrence of each dead-letter fingerprint is obtained by calculating the quotient of the number of each dead-letter fingerprint to the total number of dead letters; For each dead-letter fingerprint, when the frequency of occurrence of the dead-letter fingerprint is greater than the first threshold, the corresponding current dead-letter type is determined to be a structural anomaly type; or When the frequency of occurrence of the dead letter fingerprint is less than the second threshold, the corresponding current dead letter type is determined to be a transient anomaly type.

9. The inter-process communication control method based on historical prior constraints as described in claim 8, characterized in that, After obtaining the occurrence frequency of each dead-letter fingerprint, the method further includes: For any dead-letter fingerprint, obtain all target dead-letter messages corresponding to the arbitrary dead-letter fingerprint; When the system comprehensive load index value corresponding to more than a preset proportion of dead letters in the target dead letters is greater than a preset threshold, a resource bottleneck abnormal event is determined to have occurred. Increase the load factor of the message type corresponding to the arbitrary dead-letter fingerprint to relax the corresponding current timeout threshold and reduce the corresponding current scheduling priority.

10. The inter-process communication control method based on historical prior constraints as described in claim 3, characterized in that, The step of employing a feedback mechanism to optimize the risk prediction function based on the current dead-letter type includes: Determine the feedback score based on the current dead letter type; Obtain the learning rate and calculate the product of the learning rate and the score to obtain the adjustment step size; The weight coefficient matrix in the risk prediction function is adjusted according to the adjustment step size.

Images