Automatic management method for data center switch system

Abstract

The invention discloses an automatic management method for a data center switch system, which relates to the technical field of switches, and comprises the following steps of: calculating an error correction increment rate according to an error correction counting difference and a time difference of a line card slot position in a modularized case switch; generating a low-temperature section linear function and a high-temperature section linear function according to the air inlet temperature and the error correction increment rate of the line card slot position; and according to the low-temperature section linear function and the high-temperature section linear function, determining the target turning point temperature of the line card slot position. According to the invention, the stability and safety of the data center switch system in the automatic operation process are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of switch technology, and in particular to an automatic management method for a data center switch system. Background Technology

[0002] In recent years, with the rapid development of large-scale data centers and the widespread deployment of cloud computing, artificial intelligence, and high-frequency data analysis tasks, data center switch systems have gradually evolved towards high-density ports, high-speed interconnects, and modular architectures. Modular chassis switches typically consist of a main control unit, line cards, air ducts, and a backplane system. Their internal structure is complex, and heat distribution is highly uneven. Under long-term high-load operation of line cards, the intake air temperature and bit error rate exhibit non-linear trends, and even small changes in heat dissipation efficiency can lead to a decrease in communication stability. Modern switches generally have multi-layer scheduling and multi-level security control mechanisms. When temperature fluctuations, air duct blockages, or abnormal backplane heat conduction occur, traditional static management methods are difficult to respond in a timely manner, easily causing single-slot overheating, link degradation, or overall performance degradation.

[0003] In existing technologies, some management systems rely on periodically monitoring parameters such as temperature and fan speed for early warning or fixed threshold adjustments. However, these methods often fail to accurately identify abrupt changes in communication quality under high heat loads and lack adaptive scheduling mechanisms based on physical index correlations. This is especially true in modular chassis switches, where the airflow layout and heat dissipation interaction between line card slots are complex, and the error correction rate variations in different slots differ significantly. Traditional strategies struggle to identify potential risk points based on real-time forward error correction rates and temperature data, leading to delayed configuration updates or inconsistent security policy execution. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies where traditional strategies struggle to identify potential risk points based on real-time forward error correction rates and temperature data, leading to delayed configuration updates or inconsistent security policy execution. This invention proposes an automatic management method for data center switch systems.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution: An automatic management method for a data center switch system includes: S1. Calculate the error correction increment rate based on the error correction count difference and time difference of the line card slots in the modular chassis switch; S2. Generate a low-temperature linear function and a high-temperature linear function based on the air inlet temperature and error correction increment rate of the line card slot. S3. Determine the target inflection point temperature of the line card slot based on the linear functions of the low-temperature segment and the high-temperature segment; S4. Determine the risk index of the cable slot based on the target inflection point temperature, the inlet temperature, and the slopes of the linear functions in the low-temperature and high-temperature segments. S5. Determine the target scheduling weight of the line card slot based on the risk index, and generate change audit hashes based on the initial samples of the line card slots; S6. Under the constraints of security management, adjust the weight configuration of ports of the modular chassis switch according to the target scheduling weight and change audit hash.

[0006] Preferably, the error correction increment rate is calculated based on the error correction count difference and time difference of the line card slots in the modular chassis switch, including: Locate the cable card slots for the modular chassis switch; The error correction count difference is obtained by calculating the difference between the cumulative FEC error correction value of the line card slot at the current time and the cumulative FEC error correction value at the previous time. The time difference is calculated by comparing the current time with the previous time in the line card slot. Divide the error correction count difference by the time difference to obtain the error correction increment rate.

[0007] Preferably, generating low-temperature linear functions and high-temperature linear functions based on the inlet temperature of the line card slot and the error correction increment rate includes: Read the air inlet temperature of the cable slot; Based on the inlet temperature and the error correction increment rate, multiple initial sample pairs are generated; All inlet temperatures are discretized using a quantile grid to obtain multiple candidate breakpoints; The initial sample pairs are segmented based on the candidate breakpoints to obtain multiple low-temperature segment sample pairs and multiple high-temperature segment sample pairs. Least squares fitting is performed on all pairs of samples in the low-temperature range to obtain the linear function for the low-temperature range. Least squares fitting is performed on all high-temperature sample pairs to obtain the linear function for the high-temperature range.

[0008] Preferably, the target inflection point temperature of the wire slot is determined based on the linear functions of the low-temperature segment and the high-temperature segment, including: The error correction increment rate of the low-temperature sample pair and the fitted value of the low-temperature linear function are calculated to obtain the low-temperature residual. The sum of squares of the residuals in all low-temperature segments is obtained by summing the squares of the residuals in the low-temperature segments. The error correction increment rate of the high-temperature sample pairs and the fitted value of the high-temperature linear function are calculated to obtain the high-temperature residual. The sum of the squares of the residuals in all high-temperature segments is obtained by summing the squares of the residuals in the high-temperature segments. The sum of squared residuals in the low-temperature range is added to the sum of squared residuals in the high-temperature range to obtain the overall sum of squared residuals. The candidate breakpoint corresponding to the smallest sum of squared total residuals is determined as the target inflection point temperature of the line card slot.

[0009] Preferably, the risk index of the cable slot is determined based on the target inflection point temperature, the inlet temperature, and the slopes of the linear functions in the low-temperature and high-temperature segments, including: The temperature margin is obtained by calculating the difference between the target inflection point temperature and the current air inlet temperature of the line card slot. The slope difference is obtained by calculating the difference between the slope of the linear function in the high-temperature range and the slope of the linear function in the temperature range. The sensitivity transition amplitude is obtained by taking the maximum value of the slope difference and 0; Collect all inlet air temperatures into a temperature sample set; The interquartile range is obtained by calculating the difference between the upper and lower quartiles of the temperature sample set. The interquartile range is scaled proportionally according to a preset scaling factor to obtain the microscale term. The normalized denominator is obtained by taking the maximum value of the microscale term and the temperature margin. Divide the sensitivity jump magnitude by the normalized denominator to obtain the risk index of the line card slot.

[0010] Preferably, the target scheduling weight of the line card slot is determined based on the risk index, including: The raw score for the line card slot is calculated based on the risk index; The original scores are normalized to obtain the target scheduling weights for the line card slots.

[0011] Preferably, generating a change audit hash based on the initial sample of the line card slot includes: Record the sampling timestamps for the inlet air temperature and the error correction increment rate; Obtain the caller's identity identifier for the initial sample pair; The initial audit hash of the initial sample pair is obtained by hashing the concatenated results of the air inlet temperature, error correction increment rate, sampling timestamp and caller identity of the line card slot. Obtain the configuration change information sent from the control plane to the modular chassis switch; The concatenation result of the initial audit hash, target inflection point temperature, target scheduling weight, and configuration change content is hashed to obtain the change audit hash, so as to achieve secure management of weight configuration adjustment.

[0012] Preferably, the port weights of the modular chassis switch are adjusted according to the target scheduling weight and the change audit hash, including: Under the constraints of security management, the updated service scheduling parameters of the port are calculated based on the target scheduling weight; The result of combining the updated service scheduling parameters with the change audit hash is converted into a security configuration command for the modular chassis switch. Under the constraints of security management, the initial service scheduling parameters of the port are configured and updated according to the security configuration command.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. In this invention, by constructing an identification mechanism based on the piecewise linear relationship between the air inlet temperature of the line card slot and the FEC error correction increment rate, it is possible to detect the inflection point change of the communication performance of the slot under dynamic thermal load in real time. This effectively makes up for the technical shortcomings of traditional fixed threshold monitoring methods, which are difficult to detect the precursors of faults. By extracting the inflection point temperature through the least squares fitting algorithm and constructing a risk index system that combines the sensitivity transition amplitude and the micro-scale term, quantitative evaluation and ranking of high-risk slots are realized. This provides a scientific basis for the adaptive optimization of subsequent scheduling strategies and improves the stable operation capability of data center network equipment in non-uniform thermal environments.

[0014] 2. In this invention, a normalized calculation mechanism and a risk perception weight update method are further introduced. Combined with port scheduling scenarios and configuration distribution paths, the on-demand adjustment of port layer business scheduling parameters is realized under the automation framework. By normalizing the target scheduling weight, multiple parameters such as port scheduling priority, acceptance probability, and burst rate are driven to update in a coordinated manner, effectively balancing the contradiction between high-risk slots and their load capacity.

[0015] 3. In this invention, by constructing a change audit hash generation mechanism with initial audit hash, target inflection point temperature, target scheduling weight, and configuration change content as inputs, the integrity, traceability, and anti-tampering capability of the scheduling policy and configuration update process are ensured during transmission and execution. By encapsulating the updated service scheduling parameters into a secure configuration command with a verification signature and performing bidirectional audit verification before and after configuration distribution, a closed-loop security management system covering identification, evaluation, scheduling, and execution is constructed. This effectively avoids risks such as configuration drift, abnormal insertion, or duplicate distribution, and ensures the stability and security of the data center switch system during automated operation. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1This is a flowchart illustrating an automatic management method for a data center switch system according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Example: This example provides an automatic management method for a data center switch system. See [link to example]. Figure 1 Specifically, including: S1. Calculate the error correction increment rate based on the error correction count difference and time difference of the line card slots in the modular chassis switch; In an embodiment of the present invention, the error correction increment rate is calculated based on the error correction count difference and time difference of the line card slots in the modular chassis switch, including: Locate the cable card slots for the modular chassis switch; Specifically, modular chassis switches, as a typical representative of data center switches, feature a high-density, scalable, and multi-slot pluggable line card architecture, capable of meeting the high-throughput switching needs of east-west and north-south traffic within large-scale data centers. This type of switch is typically deployed in the core or aggregation layer, supporting multi-service scheduling, multi-level queuing, and high-speed backplane interconnection, enabling low-latency, high-bandwidth communication between massive numbers of virtual machines, containers, and compute nodes.

[0019] The error correction count difference is obtained by calculating the difference between the cumulative FEC error correction value of the line card slot at the current time and the cumulative FEC error correction value at the previous time. Specifically, firstly, the target line card slot is determined based on the location information of the chassis and backplane, and an access channel is established with the hardware counter register on that slot. Then, at the predetermined sampling time, the cumulative FEC error correction value of that slot is read as the current count, and the current sampling time is recorded synchronously. Then, the cumulative FEC error correction value and corresponding time of the same slot acquired in the previous sampling period are backtracked or cached. By subtracting the two cumulative values, the number of newly corrected errors in the sampling interval is obtained as the error correction count difference. This difference naturally eliminates the historical influence caused by the monotonic accumulation of the counter and only reflects the actual error correction workload in the most recent interval.

[0020] Specifically, a line card slot refers to the physical location unit inside a chassis-type switch used to insert line cards. It forms a fixed mechanical and thermal channel with the backplane, power supply, and air duct, and carries the physical ports and chip resources on the line card. The cumulative FEC error correction value refers to the total number of codewords or bits that the physical layer error correction circuit on the line card has successfully corrected at the current moment of device operation. This value is continuously accumulated by the chip counter and reflects the workload of forward error correction in eliminating link errors. The error correction count difference refers to the difference between the cumulative FEC error correction values ​​read at two adjacent sampling moments, which is used to indicate the number of newly corrected errors during this period.

[0021] It's important to note that FEC stands for Forward Error Correction, a coding technique used in digital communication systems to improve data transmission reliability by adding redundant checksum information at the transmitting end. This technique is implemented by transceiver chips in the physical layer of modular chassis switches, used to detect and automatically correct bit errors introduced during high-speed link transmission due to noise, crosstalk, or optical signal attenuation. When the receiving end successfully identifies and corrects the erroneous data codeword, it accumulates the corresponding number of corrections in a hardware counter, forming the cumulative FEC error correction value. This cumulative value reflects the total number of bit errors corrected during long-term link operation and is an important operational indicator for measuring link quality and reliability.

[0022] The time difference is calculated by comparing the current time with the previous time in the line card slot. Divide the error correction count difference by the time difference to obtain the error correction increment rate.

[0023] Specifically, the time difference refers to the actual elapsed time between two adjacent sampling times, used to characterize the length of the statistical interval; the error correction increment rate refers to the number of newly corrected errors per unit time, which is obtained by dividing the error correction count difference by the time difference, and is used to measure the error activity and stability margin of the current link in this slot.

[0024] Specifically, the time difference is calculated based on the system clock records at two consecutive sampling times. The result represents the actual time elapsed during signal transmission and FEC error correction within this sampling interval. According to physical laws, the rate equals the number of events occurring per unit time. When the error correction count difference represents the total number of newly corrected errors within this interval, and the time difference represents the duration of the event, the ratio of the two represents the error correction speed. The calculation of the error correction increment rate is equivalent to differentiating the cumulative FEC error correction value over time in a differential sense, thus obtaining the instantaneous change in the error correction rate. Therefore, dividing the error correction count difference by the time difference yields the error correction increment rate, which truly reflects the activity level of error correction and the trend of signal quality changes in the communication link during this time period.

[0025] Specifically, the purpose of calculating the error correction increment rate is to transform the static cumulative value of the FEC error correction in the line card slot into a dynamic indicator that reflects the trend of real-time communication quality changes. Since the cumulative FEC error correction value increases monotonically, it cannot directly reflect the sensitivity of bit error correction to changes in the operating environment. However, by calculating the increment of the error correction count within adjacent sampling periods and dividing by the time difference, the number of corrected bits per unit time can be obtained. This value exhibits a quantifiable response relationship with changes in temperature, airflow, and backplane heat dissipation, thus revealing the stability of the line card's operating status under different thermal loads. This rate indicator can accurately capture the abnormal surge in error correction load when the slot temperature transitions to the latent critical point, providing a basic input for subsequently determining the piecewise linear inflection point temperature and constructing a risk assessment model. It is a key link in realizing the automatic management and security monitoring of data center switch systems.

[0026] S2. Generate a low-temperature linear function and a high-temperature linear function based on the air inlet temperature and error correction increment rate of the line card slot. In an embodiment of the present invention, generating a low-temperature linear function and a high-temperature linear function based on the inlet temperature of the line card slot and the error correction increment rate includes: Read the air inlet temperature of the cable slot; Based on the inlet temperature and the error correction increment rate, multiple initial sample pairs are generated; Specifically, for the target line card slot in the modular chassis switch, a temperature sensor near the slot is located, and the intake air temperature data of the slot is continuously collected at multiple time points to obtain the time sequence information of temperature change in the heat dissipation environment of the slot. Based on the aforementioned collected intake air temperature data, the error correction increment rate of the slot is calculated. Then, the intake air temperature corresponding to each time point and the calculated error correction increment rate are matched one by one to generate multiple initial sample pairs.

[0027] All inlet temperatures are discretized using a quantile grid to obtain multiple candidate breakpoints; Specifically, the system first reads inlet temperature data from all line card slots of the modular chassis switch over multiple sampling periods and sorts the data in ascending order of temperature values ​​to obtain a complete temperature distribution sequence. Then, based on the sorted temperature distribution, the system uses a statistical quantile partitioning method to evenly divide the temperature intervals into several grid segments, ensuring that each segment contains approximately the same number of samples, thus guaranteeing the representativeness of each temperature interval. Next, the system extracts the corresponding temperature value at each quantile location and uses these temperature values ​​as the temperature set of candidate breakpoints. To avoid candidate points being too dense or too sparse in the boundary areas, the quantile spacing is adaptively adjusted to increase the number of candidate breakpoints in high-density temperature areas and decrease the number of candidate breakpoints in low-density areas, thereby ensuring the coverage and distribution balance of candidate breakpoints. After discretization, multiple candidate breakpoint temperature values ​​are output.

[0028] It should be noted that dividing the temperature range into several grid segments using the statistical quantile method involves first sorting all collected inlet temperature samples, and then dividing the entire temperature range into several intervals with equal probability according to the cumulative distribution probability of the samples. Each interval contains approximately the same number of temperature samples. This method differs from dividing according to fixed temperature intervals; instead, it adaptively determines the interval boundaries based on the sample distribution density. This allows for the generation of more segments in areas with denser temperature distribution and fewer segments in areas with less temperature variation. The grid segments obtained in this way more accurately reflect the statistical characteristics of the switch's operating temperature data, providing a more balanced and representative temperature distribution basis for the subsequent selection of candidate breakpoints.

[0029] Specifically, the inlet temperature of the line card slot refers to the airflow temperature measured by the temperature sensor when air enters the slot from the front of the chassis, reflecting the cooling capacity of the air duct where the slot is located; the initial sample pair refers to a set of paired observation data consisting of the inlet temperature and the corresponding error correction increment rate at the same sampling time, which are accumulated in chronological order to form a data sequence for modeling; quantile grid discretization refers to selecting a set of representative temperature values ​​based on several quantiles of the historical temperature distribution of the slot to form a scanning grid, thereby transforming the continuous temperature axis into a finite number of nodes to be evaluated; candidate breakpoints are multiple representative temperature nodes generated based on the discrete values ​​of the inlet temperature distribution, each node corresponding to a possible linear turning point, used to divide the low-temperature segment and the high-temperature segment of the temperature range.

[0030] The initial sample pairs are segmented based on the candidate breakpoints to obtain multiple low-temperature segment sample pairs and multiple high-temperature segment sample pairs. Specifically, the low-temperature sample pair is the set of observation data for all inlet air temperatures not higher than the candidate breakpoint, reflecting the error correction behavior of the equipment under low thermal load conditions; the high-temperature sample pair is the set of observation data for all inlet air temperatures higher than the candidate breakpoint, used to characterize the error correction features of the equipment under high thermal load conditions.

[0031] Specifically, for each candidate breakpoint temperature, all samples in the initial sample set are traversed. Each sample contains the inlet temperature and error correction increment rate corresponding to the sampling time. The sample is assigned to a specific interval based on its temperature. If the inlet temperature of a sample is not higher than the current candidate breakpoint temperature, the sample is assigned to the low-temperature interval sample set. If the inlet temperature of a sample is higher than the current candidate breakpoint temperature, the sample is assigned to the high-temperature interval sample set. During the partitioning process, the time order and index of samples in the same slot are kept consistent to prevent sample mismatch or cross-regional partitioning. After partitioning, the number of low-temperature interval samples and the number of high-temperature interval samples corresponding to the candidate breakpoint are recorded respectively, and the coverage ratio of each interval sample is calculated. If the number of samples in any temperature interval is insufficient to support linear fitting, the candidate breakpoint is automatically discarded, and breakpoints that meet the statistical requirements are retained for the next step of calculation. This process is repeated cyclically at all candidate breakpoint temperatures, ultimately forming a set of filtered low-temperature interval sample sets and high-temperature interval sample sets.

[0032] Least squares fitting is performed on all pairs of samples in the low-temperature range to obtain the linear function for the low-temperature range. Least squares fitting is performed on all high-temperature sample pairs to obtain the linear function for the high-temperature range.

[0033] Specifically, least squares fitting is a regression calculation process based on statistical principles. It determines the optimal linear function coefficients by minimizing the sum of squared residuals between the sample observations and the model predictions, thereby obtaining linear functions for the low-temperature segment and the high-temperature segment. These functions can approximate the changes in temperature and error correction increment rate in different thermal ranges, providing a quantitative basis for subsequent identification of temperature-related inflection points.

[0034] Specifically, the low-temperature linear function refers to a linear expression obtained by fitting the inlet temperature as the independent variable and the error correction increment rate as the dependent variable within the temperature range where the inlet temperature does not exceed the candidate breakpoint, using the least squares method. It is used to describe the approximate law of the error correction rate of the modular chassis switch line card changing with temperature under low heat load conditions. This function reflects the stable operating characteristics of the device when the cooling is sufficient and the bit error rate is low. The high-temperature linear function refers to a linear expression obtained by fitting the inlet temperature with the least squares method within the temperature range after the inlet temperature exceeds the candidate breakpoint. It is used to characterize the response relationship of the line card's error correction rate increasing with temperature under high heat load conditions. This function reflects the performance degradation trend of the device when the heat dissipation is insufficient and the bit error rate increases. Together, the two constitute a piecewise linear model of the relationship between temperature and error correction rate, which is used to identify the characteristic points where temperature changes cause a turning point in link stability.

[0035] Specifically, the process first reads the low-temperature segment sample set generated according to candidate breakpoints. The sample consists of the inlet temperature at the same moment and the corresponding error correction increment rate. After missing data removal and time sequence verification, a linear relationship hypothesis is established with the inlet temperature as the independent variable and the error correction increment rate as the dependent variable. A data matrix and observation vector composed of constant terms and temperature terms are constructed. The sum of squared residuals between the observed values ​​and the linear expression is minimized using the ordinary least squares method to obtain the slope and intercept of the low-temperature segment linear function. At the same time, the sum of squared residuals and the sample coverage ratio are calculated as fitting quality indicators and persisted. Subsequently, the same process is performed on the high-temperature segment sample set, including sample validity check, linear model setting, least squares solution and quality indicator calculation, to obtain the slope and intercept of the high-temperature segment linear function and its corresponding fitting quality. Finally, the applicable linear functions for the two online temperature ranges and the corresponding quality indicators are output.

[0036] S3. Determine the target inflection point temperature of the line card slot based on the linear functions of the low-temperature segment and the high-temperature segment; In an embodiment of the present invention, determining the target inflection point temperature of the wire slot based on the low-temperature segment linear function and the high-temperature segment linear function includes: The error correction increment rate of the low-temperature sample pair and the fitted value of the low-temperature linear function are calculated to obtain the low-temperature residual. Specifically, the low-temperature residual refers to the difference between the actual error correction increment rate of the line card slot and the fitted rate calculated based on the low-temperature linear function under low-temperature operating conditions. This difference reflects the degree of deviation of the communication bit error rate of the equipment from the temperature change in a low-temperature environment. The fitted value of the low-temperature linear function refers to the calculation result of the error correction increment rate predicted by the low-temperature linear function established by the least squares method within the low-temperature operating range. The fitted value represents the trend of error correction rate change that the equipment should theoretically exhibit under the current temperature conditions, reflecting the linear response relationship between temperature and bit error rate in the low-temperature range.

[0037] The sum of squares of the residuals in all low-temperature segments is obtained by summing the squares of the residuals in the low-temperature segments. The error correction increment rate of the high-temperature sample pairs and the fitted value of the high-temperature linear function are calculated to obtain the high-temperature residual. Specifically, the sum of squares of the low-temperature residuals is the total sum of the squares of all low-temperature residuals. It measures the overall approximation of the linear fitting model to the actual measurement data within the low-temperature range. The smaller the sum of squares, the higher the model's fitting accuracy for the low-temperature samples. The high-temperature residual refers to the deviation between the actual error correction increment rate and the predicted rate of the high-temperature linear function after the temperature rises to the high-temperature range. This deviation reveals the change in bit error rate caused by airflow resistance, decreased heat dissipation efficiency, or signal attenuation at high temperatures. By calculating the sum of squares of the high-temperature residuals, the fitting error level of the model in the high-temperature range can be evaluated. The residuals and their sums of squares in both the low-temperature and high-temperature ranges reflect the impact of temperature changes on the stability of the communication link, providing a quantitative basis for determining the target inflection point temperature of the line card slot.

[0038] Specifically, the calculation of residuals in the low-temperature and high-temperature ranges is based on the principles of physical measurement error and function approximation. The core idea is to fit a linear function to the relationship between temperature and the error correction increment rate, and then measure the accuracy of the model by comparing the difference between the model's predicted values ​​and the actual observed values. According to basic physical formulas, the actual output of any measurement system can be expressed as the sum of the ideal response and the error components. The calculated value obtained from the linear fitting represents the theoretical response of the system under ideal steady-state conditions, while the actually observed error correction increment rate is the true output affected by factors such as airflow disturbances, heat dissipation efficiency, and electrical noise. Therefore, the difference between the two reflects the dynamic deviation of the equipment's operation from steady-state conditions; this deviation is the residual. In the low-temperature range, this residual reflects the discreteness of error correction when the equipment is sufficiently cooled; in the high-temperature range, this residual reveals the nonlinear fluctuations in the error correction performance of the equipment when the thermal load increases. The residuals obtained through this calculation method not only conform to the definition of measurement error and model approximation principles but also accurately reflect the communication link stability shift caused by temperature changes.

[0039] The sum of the squares of the residuals in all high-temperature segments is obtained by summing the squares of the residuals in the high-temperature segments. The sum of squared residuals in the low-temperature range is added to the sum of squared residuals in the high-temperature range to obtain the overall sum of squared residuals. Specifically, the high-temperature residual sum of squares refers to the sum of the squares of the high-temperature residual values ​​of each sample point within the high-temperature operating range. This value is used to measure the magnitude of the fitting error of the linear model to the actual error correction rate data under high-temperature conditions, reflecting the degree of deviation of the communication stability when the heat dissipation is limited or the heat load increases. The low-temperature residual sum of squares is the sum of the squares of all low-temperature residual values ​​within the low-temperature operating range. It is used to characterize the difference between the model fitting and the actual operation under conditions of sufficient cooling and few bit errors, representing the error accumulation of the system under stable operating conditions. The overall residual sum of squares is the sum of the low-temperature residual sum of squares and the high-temperature residual sum of squares, reflecting the overall approximation of the linear fitting model to the actual samples within the entire temperature range. The smaller this value, the more consistent the model is with the actual operating trend.

[0040] The candidate breakpoint corresponding to the smallest sum of squared total residuals is determined as the target inflection point temperature of the line card slot.

[0041] Specifically, the target inflection point temperature refers to the boundary temperature value in the relationship curve between the temperature of the line card slot and the error correction increment rate that minimizes the sum of the linear fitting errors of the low-temperature and high-temperature segments. This temperature point represents the critical transition position of the equipment's operating state from a stable stage to a performance degradation stage. When the inlet temperature is below this temperature, the error correction increment rate changes relatively slowly with temperature, indicating that the equipment is in a normal heat dissipation and stable communication state. However, when the temperature exceeds this point, the error correction increment rate begins to rise significantly, reflecting an increase in link errors, a decrease in heat dissipation efficiency, or damage to signal integrity. This temperature value physically reveals the boundary conditions for the transition of the modular chassis switch line card from the thermally stable zone to the risk zone, and is a key parameter for judging the thermal health status of the equipment and triggering automatic management and scheduling.

[0042] Specifically, the determination of the target inflection point temperature for the line card slot is based on thermodynamic changes and the least squares error principle. As the inlet temperature increases, the power consumption of the internal circuitry and optical modules causes heat accumulation, leading to signal attenuation and an increase in the bit error rate, manifested as a nonlinear increase in the error correction increment rate. When the temperature is in the low-temperature range, the bit error correction process is less affected by noise and slight thermal disturbances, and the system response is approximately linear. However, as the temperature rises further, the heat conduction efficiency decreases, and a temperature gradient forms locally in the heat dissipation duct, causing a rapid increase in the error correction rate. By superimposing the sum of squared residuals from the low-temperature and high-temperature segments and taking their minimum value, we are essentially looking for the optimal boundary point in the temperature-error correction rate relationship curve, minimizing the energy deviation of the fitted model across the overall temperature range. The temperature value corresponding to this minimum error precisely reflects the critical condition for the system to transition from a thermal equilibrium state to a thermally unstable state. Therefore, the temperature represented by this point is determined as the target inflection point temperature for the line card slot, used to define the physical boundary of equipment performance changes and as a key judgment basis for automatic management and scheduling.

[0043] S4. Determine the risk index of the cable slot based on the target inflection point temperature, the inlet temperature, and the slopes of the linear functions in the low-temperature and high-temperature segments. In embodiments of the present invention, the risk index of the cable slot is determined based on the target inflection point temperature, the inlet temperature, and the slopes of the low-temperature linear function and the high-temperature linear function, including: The temperature margin is obtained by calculating the difference between the target inflection point temperature and the current air inlet temperature of the line card slot. The slope difference is obtained by calculating the difference between the slope of the linear function in the high-temperature range and the slope of the linear function in the temperature range. The sensitivity transition amplitude is obtained by taking the maximum value of the slope difference and 0; Specifically, temperature margin refers to the difference between the current air inlet temperature of the line card slot and its target inflection point temperature. This difference characterizes the safety margin of the equipment's current operating state from the thermal instability critical point. A larger temperature margin indicates that the system is in a relatively safe thermal equilibrium range, while a smaller margin indicates that the system is approaching or has entered a risk zone. Slope difference refers to the difference between the slope of the linear function in the high-temperature segment and the slope of the linear function in the low-temperature segment. It represents the difference in the response strength of the error correction increment rate as the temperature rises. A larger slope difference indicates that the temperature change has a more significant impact on the stability of the communication link. Sensitivity transition amplitude is a quantitative indicator obtained by comparing the slope difference with zero and taking the maximum value. It is used to measure the degree of abrupt change in the temperature response characteristics of the equipment. This indicator comprehensively reflects the severity of the system's transition from stable operation to thermal stress state and is an important basis for judging whether the air duct and line card coordination in modular chassis switches are unbalanced.

[0044] Specifically, the sensitivity transition amplitude is calculated based on the nonlinear variation of heat conduction and signal response. When the line card slot is in different temperature ranges, the relationship between the inlet air temperature and the error correction increment rate exhibits different linear responses. In the low-temperature range, the system is in thermal equilibrium, and the current carrying and heat dissipation processes within the device are relatively stable, resulting in a slow change in the error correction increment rate with increasing temperature. However, in the high-temperature range, the increase in temperature causes an increase in device resistance, enhanced noise power, and unstable signal modulation, leading to a rapid increase in the error correction increment rate. At this point, the slope of the linear function represents the system's response strength to temperature changes. By calculating the difference between the slopes of the linear function in the high-temperature and low-temperature ranges, the degree of abrupt change in the system's response when crossing thermally stable and unstable regions can be quantified. Furthermore, maximizing the slope difference against zero ensures that only the sensitivity enhancement caused by temperature increases is extracted, thus avoiding interference from negative errors in the evaluation results. The final sensitivity transition amplitude reflects the drastic change in system performance under thermal disturbance conditions. Its physical meaning corresponds to the transition strength of the device's thermal dynamic response and can be used to characterize the stability boundary of modular chassis switches under different thermal loads.

[0045] Collect all inlet air temperatures into a temperature sample set; The interquartile range is obtained by calculating the difference between the upper and lower quartiles of the temperature sample set. Specifically, the temperature sample set is a collection of inlet temperature data from all line card slots, reflecting the overall thermal distribution of the system. The upper quartile refers to the temperature value at the 75th percentile of all sample values ​​after arranging the temperature sample set from smallest to largest, reflecting the concentration trend of high-temperature components. The lower quartile refers to the temperature value at the 25th percentile of the sample set, representing the concentration trend of low-temperature components. The upper and lower quartiles together characterize the middle range of the system's temperature distribution, that is, the stable range in which the operating temperature of most line card slots is located. The interquartile range is the difference between the upper and lower quartiles, representing the overall dispersion and fluctuation of the temperature samples. A larger value indicates a more significant temperature difference between slots in the system and a more uneven heat dissipation. A smaller interquartile range indicates a more concentrated temperature distribution and better thermal balance within the chassis. This indicator reflects the uniformity of thermal distribution in the data center switch system through statistical regularity and is an important parameter for measuring temperature control stability.

[0046] The interquartile range is scaled proportionally according to a preset scaling factor to obtain the microscale term. Specifically, the calculated interquartile range is multiplied by a preset scaling factor to obtain a microscale term, thereby achieving proportional scaling of temperature fluctuation scale.

[0047] Specifically, the preset scaling factor is a dimensionless adjustment parameter manually set during temperature distribution analysis. It maps macroscopic temperature fluctuations to a microscopic scale to achieve standardized scaling of temperature differences between different slots. This scaling factor is set by comprehensively considering factors such as chassis airflow design, heat transfer efficiency, and ambient temperature fluctuations, ensuring consistent sensitivity in the calculation process across different equipment and operating environments. The microscale term is a correction obtained by multiplying the interquartile range by the preset scaling factor. It characterizes the microscopic temperature change scale of the system under local thermal disturbances. This term can capture subtle thermal anomalies beyond macroscopic temperature changes, providing a basis for subsequent temperature risk assessment and scheduling weight allocation. By introducing the microscale term, the accuracy of the response to instantaneous temperature differences and local thermal accumulation effects can be improved while maintaining the accuracy of overall thermal trend analysis, thereby enhancing the security management and adaptive capabilities of the data center switch system's automatic management process.

[0048] Specifically, the calculation of the micro-scale term is based on the principles of statistical thermodynamics and the stability analysis of temperature distribution. The airflow and heat conduction processes inside a modular chassis switch exhibit complex local disturbances. The inlet temperature of different line card slots is discretely distributed due to factors such as duct structure, air pressure distribution, and power consumption differences. By statistically calculating the inlet temperature samples of all slots and using the difference between the upper and lower quartiles, the dispersion of the overall temperature distribution, i.e., the interquartile range, can be obtained. This value reflects the amplitude of temperature fluctuations at the macroscopic level of the system. To refine the handling of local temperature fluctuations in the automatic management algorithm, this macroscopic feature needs to be scaled down to a microscopic scale to reflect subtle thermal disturbances. By multiplying the interquartile range by a preset scaling factor, the overall temperature fluctuation can be mapped to a scale quantity of local thermal imbalances, thereby constructing a temperature perturbation correction term with adaptive sensitivity.

[0049] It should be noted that the generation steps of the preset proportional coefficient include establishing a multi-factor parameter model based on the equipment structural characteristics and operating environment, and determining the proportional weights by statistically analyzing historical operating data and thermal distribution characteristics. First, temperature distribution data of the modular chassis switch under different loads and airflow conditions are collected to obtain the inlet temperature fluctuation range of each line card slot and the corresponding FEC error correction increment rate change amplitude. Second, a correlation analysis is performed on the temperature fluctuation amplitude and the bit error rate change rate to extract the sensitivity coefficient of the impact of temperature disturbance on communication performance. Then, combined with the overall chassis heat dissipation design parameters, airflow distribution characteristics, and average power consumption density, a basic proportional weight is formed through normalization. This weight is then weighted and averaged with the temperature risk tolerance parameter in the system security management strategy to obtain a proportional benchmark applicable across the entire rack scale. Finally, the proportional benchmark is dynamically corrected so that it can automatically adjust according to changes in real-time ambient temperature and airflow efficiency, thereby generating an adaptive preset proportional coefficient, providing a stable and universally applicable quantitative basis for subsequent temperature interquartile range scaling and micro-scale term calculations.

[0050] The normalized denominator is obtained by taking the maximum value of the microscale term and the temperature margin. Divide the sensitivity jump magnitude by the normalized denominator to obtain the risk index of the line card slot.

[0051] Specifically, the normalized denominator is the baseline scale obtained by taking the maximum value of the micro-scale term and the temperature margin. Its function is to provide a stable and non-zero standardized scale when the equipment temperature is close to the critical point, thereby avoiding divergence in risk calculation. The risk index is a dimensionless measure obtained by dividing the sensitivity jump amplitude by the normalized denominator. Its magnitude is affected by both the intensity of response mutation and the safety margin. The larger the value, the closer the line card slot is to the thermal risk boundary and the more susceptible the communication quality is to temperature disturbances.

[0052] Specifically, the risk index for line card slots is calculated based on the coupling relationship between system thermal balance and transmission performance changes. Temperature margin represents the safe distance between the current operating temperature and the thermal stability threshold, while the micro-scale term reflects the typical fluctuation range of local temperature differences within the chassis. By taking the maximum of both and using it as the normalized denominator, a unified temperature scale benchmark is ensured for the risk assessment calculation, preventing over-amplification or distortion of the results when equipment temperature approaches the critical point or ambient temperature fluctuates drastically. The sensitivity transition amplitude corresponds to the maximum positive value of the difference in linear response slope between high and low temperature regions, physically representing the intensity of the change in signal error correction rate caused by thermal disturbance. When the system is in stable thermal balance, the slope difference approaches zero, while it increases significantly under sudden changes in thermal load or uneven cooling. The result of dividing the sensitivity transition amplitude by the normalized denominator, i.e., the risk index, reflects the relative impact of temperature disturbance on communication performance. Its value simultaneously embodies the comprehensive characteristics of thermal safety margin and signal stability, thus enabling the quantification and dynamic identification of potential thermal risks in each line card slot within the automatic management system.

[0053] In general, the purpose of calculating the risk index for line card slots is to achieve coordinated control of security management and operational scheduling during the automated management of data center switch systems. During long-term operation, the response relationship between the inlet temperature and error correction increment rate (FEC) of line card slots in modular chassis switches dynamically changes with environmental factors, power consumption, and airflow disturbances. When the temperature exceeds a certain inflection point, the FEC error correction rate increases sharply, indicating that the equipment has entered a thermally unstable state. By constructing a risk index, this implicit coupling process between temperature and performance can be quantified, enabling the system to automatically identify potential thermal risks without relying on manual monitoring. The risk index comprehensively considers temperature margin, micro-scale terms, and sensitivity transition amplitude, reflecting both the distance of the system from the safe hot zone and the degree of impact of temperature disturbances on link quality. This indicator plays an early warning and regulatory role in the security management layer, serving as a dynamic input for configuration adjustments and airflow optimization, ensuring the long-term reliability and security of data center switches under high-density multi-line-card operation conditions.

[0054] S5. Determine the target scheduling weight of the line card slot based on the risk index, and generate change audit hashes based on the initial samples of the line card slots; In an embodiment of the present invention, determining the target scheduling weight of the line card slot based on the risk index includes: The raw score for each line card slot is calculated based on the risk index. The formula for calculating the raw score is as follows: In the formula, It is the first Risk index of each line card slot It is the first The original score for each line card slot; The original scores are normalized to obtain the target scheduling weights for the line card slots.

[0055] Specifically, the original scores of all slots are aggregated and their sum is calculated to construct the normalized denominator benchmark; then, the original score of each slot is divided by the sum of the original scores of all slots to obtain the target scheduling weight.

[0056] Specifically, the raw score of a line card slot characterizes its basic security level within the risk indicator system. It is obtained by inversely mapping the risk index to a comparable normalized value. The calculation uses a summation of the inverse, ensuring that the raw score decreases as the risk index of the line card slot increases, thus reflecting the principle that high-risk slots should have lower priority or scheduling weight in the overall system. The raw score essentially reflects the slot's comprehensive performance in terms of thermal stability, error correction capability, and operational safety margin. The target scheduling weight is a proportional factor obtained by normalizing the raw score, used to allocate the operational priority of different line card slots during system scheduling and resource allocation. Through normalization, the total weight of all line card slots remains at a uniform standard, enabling the system to achieve overall balanced thermal load and service scheduling under different slot risk differences, thereby ensuring the stable operation of network switches under security constraints.

[0057] Specifically, the target scheduling weight is determined in two steps: a monotonically inverse mapping from risk to weight and total normalization. First, the original score is obtained by adding the inverse of the risk index, so that the higher the risk, the lower the score; when the risk is 0, the score equals 1; and when the risk approaches infinity, the score approaches 0. This normalizes the risk amount in the infinite range to between zero and one, maintaining boundedness and numerical stability. This inverse mapping ensures that scheduling priority is inversely proportional to safety risk, which conforms to the control principle that high-risk areas should be load-reduced. Then, the original scores of all slots are summed as a benchmark, and the original score of each slot is divided by this benchmark to obtain a proportion value, so that the sum of the proportions of all slots equals 1. This transforms relative safety into a directly executable resource allocation share. The value obtained after this mapping and normalization is comparable, additive, and conserved. It can directly transform the risk assessment result into a dispatchable scheduling parameter without relying on manual thresholds. Therefore, this result is the target scheduling weight of the line card slot.

[0058] In general, the fundamental purpose of calculating the target scheduling weight for line card slots is to achieve dynamic adaptive and risk-constrained control of data center switch resource allocation within a security management framework. Due to the influence of airflow distribution, heat dissipation structure, and environmental disturbances during operation, different slots exhibit significant differences in their inlet temperature and error correction increment rate variations. These differences directly reflect the differences in thermal risk and link stability of the slots. By calculating a risk index and further converting it into a normalized target scheduling weight, these differences can be quantified into a system-executable scheduling ratio. This allows high-risk slots to receive lower operating priorities, while low-risk slots maintain or increase their service carrying ratio. This approach not only ensures a balance between internal switch thermal load and computational tasks but also achieves risk-driven secure resource allocation. The target scheduling weight is used to constrain configuration updates and task scheduling processes, enabling the system to proactively adjust resource allocation strategies when potential abnormal temperature rises, abnormal error correction increments, or heat dissipation imbalances occur. This avoids system-level failures caused by localized overheating or link instability, thereby improving the overall security of network equipment.

[0059] In an embodiment of the present invention, generating a change audit hash based on an initial sample of the line card slot includes: Record the sampling timestamps for the inlet air temperature and the error correction increment rate; Obtain the caller's identity identifier for the initial sample pair; Specifically, the sampling timestamp refers to the actual time point recorded by a unified clock when reading the set of physical quantities, namely the inlet temperature and the error correction increment rate. It is used to determine the actual occurrence time corresponding to the set of observations and to provide a benchmark for subsequent differential calculations and timing alignment. The caller identity identifier refers to the specific entity that triggered this sampling and data reporting. It can be an entity represented by a sampling task, management process, or controlled interface running in the control plane. It is used to bind each set of initial sample pairs to its source, thereby realizing data source confirmation, access auditing, and non-repudiation tracing in security management.

[0060] The initial audit hash of the initial sample pair is obtained by hashing the concatenated results of the air inlet temperature, error correction increment rate, sampling timestamp and caller identity of the line card slot. Obtain the configuration change information sent from the control plane to the modular chassis switch; Specifically, the initial audit hash refers to a fixed-length digest value generated by concatenating basic operational data such as the air inlet temperature, error correction increment rate, sampling timestamp, and caller identification of the line card slot, using a hash algorithm. This digest uniquely identifies the sample data status at that moment. Essentially, it's an integrity verification identifier based on a cryptographic hash function, ensuring that any subsequent configuration adjustments or status backtracking can verify whether the original data has been tampered with or forged. It's a key element for achieving system security auditing and trusted data transmission. Configuration change content refers to the set of parameter update instructions issued from the control plane to the modular chassis switch, including changes to port weights, scheduling priorities, resource allocation ratios, etc., reflecting the dynamic adjustment of the system's operating status.

[0061] The concatenation result of the initial audit hash, target inflection point temperature, target scheduling weight, and configuration change content is hashed to obtain the change audit hash, so as to achieve secure management of weight configuration adjustment.

[0062] Specifically, to ensure the security and traceability of system configuration adjustments, multiple key operating parameters need to be combined to generate security verification information. Specifically, the initial audit hash, the target inflection point temperature of the line card slot, the target scheduling weight, and the configuration changes issued by the control plane are first concatenated to form a complete change context record. Then, a hash calculation is performed on the concatenated result to generate a unique corresponding change audit hash. This change audit hash, as part of the security management mechanism, can verify the integrity and consistency of the configuration issuance process and enable secure tracking of configuration reset and scheduling adjustment operations, preventing risks caused by instruction tampering or unauthorized operations, thereby ensuring the security, controllability, and compliant management of the switch system in dynamic scheduling.

[0063] Specifically, the purpose of generating change audit hashes is to achieve integrity protection and traceability control over the entire configuration change process during the automated management of data center switches through security management methods. Because modular chassis switches frequently need to adjust port scheduling weights, airflow parameters, or load distribution strategies during dynamic operation, these operations, if maliciously tampered with, mistakenly issued, or repeatedly executed, may cause security risks such as network outages or thermal imbalances. By concatenating the initial audit hash, target inflection point temperature, target scheduling weight, and configuration change content to calculate the change audit hash, a unique encrypted digest can be generated for each change, proving the authenticity of the association between the configuration command and the operating status. This hash value can not only verify the legality of the operation in subsequent system checks but also track the source and time of each change in security audits, thus forming a tamper-resistant secure link to ensure that data center switches remain within a trusted and controlled security management system during automated scheduling and configuration updates.

[0064] S6. Under the constraints of security management, adjust the weight configuration of ports of the modular chassis switch according to the target scheduling weight and change audit hash.

[0065] In an embodiment of the present invention, the port weight configuration of the modular chassis switch is adjusted according to the target scheduling weight and the change audit hash, including: Under the constraints of security management, the updated service scheduling parameters of the port are calculated based on the target scheduling weight; Specifically, firstly, the line card slot corresponding to the target scheduling weight is determined, and all physical ports carried by that line card slot are identified, establishing a one-to-one correspondence between the line card slot and the physical ports to ensure that subsequent parameter calculations can be accurately associated with specific ports. Under the constraints of security management, the previously generated change audit hash is first called to verify the legality and integrity of the target scheduling weight, confirming that the target scheduling weight is calculated based on an effective risk index and has not been tampered with, thus meeting security management requirements. Subsequently, according to the service scheduling rules of the modular chassis switch, for port inbound acceptance scenarios, the target scheduling weight is converted into the port inbound acceptance probability. The higher the target scheduling weight, the greater the inbound acceptance probability of the corresponding port. Simultaneously, each parameter is verified to ensure it conforms to the switch hardware configuration specifications, such as whether the weight value is within the effective range supported by the device and whether the probability value is within a reasonable range of 0 to 1. Finally, all calculation results are integrated to form the updated service scheduling parameter set for the port corresponding to the line card slot, and each parameter is associated with the corresponding target scheduling weight and security audit identifier to ensure that the parameter calculation process is traceable and complies with security management constraints.

[0066] Specifically, in port inbound admission scenarios, the target scheduling weight can be used as a unified control gain to simultaneously calculate multiple admission parameters: First, the port inbound admission probability is obtained through a linear mapping of the target scheduling weight, with higher weights resulting in higher admission probabilities; then, using the physical port rate or historical stable throughput as a baseline rate, the target scheduling weight is applied proportionally to the baseline rate to obtain the average admission rate; next, the baseline token count is scaled with the same weight to obtain the admission burst quota, thereby determining the average token bucket replenishment rate and available burst capacity; simultaneously, the congestion labeling gain is scaled with the target scheduling weight to obtain the explicit congestion notification label. The system calculates the intensity of congestion, allowing high-weight ports to be marked later under mild congestion. Further, it uses the target scheduling weight to perform linear interpolation between the minimum and maximum queue levels to obtain the upper and lower thresholds and drop slope for random early drops, giving high-weight ports a wider acceptable queue range. Finally, it scales the acceptance time window and backoff factor proportionally using the target scheduling weight to obtain the evaluation period and backoff speed for acceptance decisions. This forms a consistent parameter set consisting of acceptance probability, average rate, burst limit, congestion marking gain, early drop threshold, and backoff parameters, ensuring that port inbound acceptance aligns with system scheduling goals and meets safety management constraints.

[0067] The result of combining the updated service scheduling parameters with the change audit hash is converted into a security configuration command for the modular chassis switch. Specifically, in the process of generating secure configuration commands, the updated service scheduling parameters are first formatted under the constraints of the security management module to ensure that the parameter units, numerical precision, and scheduling levels conform to the configuration specifications of the modular chassis switch. Then, the change audit hash is invoked from the security control plane. The hash value is used to concatenate and combine the currently updated service scheduling parameters to generate a unique associated string. A digest operation is then performed using an encrypted hash algorithm to obtain an irreversible secure signature identifier, ensuring the authenticity and integrity of the scheduling parameter set. Next, based on the chassis topology information, the system expands the signed parameter data according to the slot and port mapping relationship, generating corresponding port configuration command templates. Based on the execution priority, dependency order, and rollback strategy of the configuration commands, the command content is transformed into a secure configuration command set. During the formation of this command set, a timestamp, caller identity, and operation credentials are simultaneously appended for subsequent security audit tracing and verification. Finally, the generated configuration command is validated and signed in the security management module. Once validated, it is sent to the control plane of the modular chassis switch. This achieves encrypted protection and trusted execution of the configuration distribution process, ensuring that the binding relationship between updating service scheduling parameters and changing audit hashes is complete, secure, and traceable.

[0068] Under the constraints of security management, the initial service scheduling parameters of the port are configured and updated according to the security configuration command.

[0069] Specifically, updating service scheduling parameters refers to a new set of parameters that are dynamically adjusted based on the calculated target scheduling weights, including the service resource allocation ratio, bandwidth scheduling priority, and load balancing weights for each port or line card slot. This new set of parameters guides the real-time scheduling behavior of the switch. A security configuration command is a configuration issuance instruction with tamper-proof characteristics generated by the security management module after verifying the change audit hash. This command, combined with encryption signatures or hash verification mechanisms, ensures the integrity and trustworthiness of the configuration content during transmission and execution. Initial service scheduling parameters refer to the scheduling parameter values ​​previously stored in each port or slot before the system performs the update operation. These parameters reflect the service allocation strategy of the device in its previous stable operating state.

[0070] Specifically, the process of configuring and updating the initial service scheduling parameters of a port under the constraints of security management includes the following steps: First, the security management module performs secondary verification on the security configuration command to be executed, confirming that the source identity, digital signature, timestamp, and change audit hash of the command are consistent with those recorded in the audit log to prevent unauthorized commands or man-in-the-middle attacks; then, the system freezes the current scheduling parameter set of the target port, identifies it as the initial service scheduling parameters, and generates a backup in the secure storage area to support fault rollback; next, the control plane parses the updated service scheduling parameter content in the security configuration command, loads it item by item according to the port level, and verifies the consistency of the parameters with port attributes, bandwidth constraints, and duct power consumption limits to ensure that the new configuration will not cause resource issues. Source conflicts or thermal imbalances are identified; then, the switch's command execution engine writes the updated parameters into the port's run table entry, and synchronously updates the scheduling queue, traffic shaping, priority mapping, and packet scheduling weight, enabling the port to complete parameter hot replacement without interrupting communication; after the configuration application is completed, the security management module immediately generates a new status hash, compares it with the original change audit hash to verify configuration consistency, if they match, writes this update record to the audit log and releases the port lock state; finally, it outputs the configuration success status information and marks the updated service scheduling parameters as the currently valid parameters, realizing port-level secure configuration updates, thereby ensuring the configuration reliability, traceability, and operational security of the modular chassis switch under automated management.

[0071] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An automatic management method for a data center switch system, characterized in that, Includes the following steps: S1. Calculate the error correction increment rate based on the error correction count difference and time difference of the line card slots in the modular chassis switch; S2. Generate a low-temperature linear function and a high-temperature linear function based on the air inlet temperature and error correction increment rate of the line card slot. S3. Determine the target inflection point temperature of the line card slot based on the linear functions of the low-temperature segment and the high-temperature segment; S4. Determine the risk index of the cable slot based on the target inflection point temperature, the inlet temperature, and the slopes of the linear functions in the low-temperature and high-temperature segments. S5. Determine the target scheduling weight of the line card slot based on the risk index, and generate change audit hashes based on the initial samples of the line card slots; S6. Under the constraints of security management, adjust the weight configuration of ports of the modular chassis switch according to the target scheduling weight and change audit hash.

2. The automatic management method for a data center switch system according to claim 1, characterized in that, Based on the error correction count difference and time difference of the line card slots in the modular chassis switch, calculate the error correction increment rate, including: Locate the cable card slots for the modular chassis switch; The error correction count difference is obtained by calculating the difference between the cumulative FEC error correction value of the line card slot at the current time and the cumulative FEC error correction value at the previous time. The time difference is calculated by comparing the current time with the previous time in the line card slot. Divide the error correction count difference by the time difference to obtain the error correction increment rate.

3. The automatic management method for a data center switch system according to claim 1, characterized in that, Based on the inlet temperature of the line card slot and the error correction increment rate, generate low-temperature and high-temperature linear functions, including: Read the air inlet temperature of the cable slot; Based on the inlet temperature and the error correction increment rate, multiple initial sample pairs are generated; All inlet temperatures are discretized using a quantile grid to obtain multiple candidate breakpoints; The initial sample pairs are segmented based on the candidate breakpoints to obtain multiple low-temperature segment sample pairs and multiple high-temperature segment sample pairs. Least squares fitting is performed on all pairs of samples in the low-temperature range to obtain the linear function for the low-temperature range. Least squares fitting is performed on all high-temperature sample pairs to obtain the linear function for the high-temperature range.

4. The automatic management method for a data center switch system according to claim 1, characterized in that, Based on the linear functions of the low-temperature and high-temperature segments, the target inflection point temperature of the wire clamp slot is determined, including: The error correction increment rate of the low-temperature sample pair and the fitted value of the low-temperature linear function are calculated to obtain the low-temperature residual. The sum of squares of the residuals in all low-temperature segments is obtained by summing the squares of the residuals in the low-temperature segments. The error correction increment rate of the high-temperature sample pairs and the fitted value of the high-temperature linear function are calculated to obtain the high-temperature residual. The sum of the squares of the residuals in all high-temperature segments is obtained by summing the squares of the residuals in the high-temperature segments. The sum of squares of the residuals in the low-temperature range is added to the sum of squares of the residuals in the high-temperature range to obtain the overall sum of squares of the residuals. The candidate breakpoint corresponding to the smallest sum of squared residuals is determined as the target inflection point temperature of the line card slot.

5. The automatic management method for a data center switch system according to claim 1, characterized in that, Based on the target inflection point temperature, inlet temperature, and the slopes of the linear functions in the low-temperature and high-temperature segments, the risk index of the cable slot is determined, including: The temperature margin is obtained by calculating the difference between the target inflection point temperature and the current air inlet temperature of the line card slot. The slope difference is obtained by calculating the difference between the slope of the linear function in the high-temperature range and the slope of the linear function in the temperature range. The sensitivity transition amplitude is obtained by taking the maximum value of the slope difference and 0; Collect all inlet air temperatures into a temperature sample set; The interquartile range is obtained by calculating the difference between the upper and lower quartiles of the temperature sample set. The interquartile range is scaled proportionally according to a preset scaling factor to obtain the microscale term. The normalized denominator is obtained by taking the maximum value of the microscale term and the temperature margin. Divide the sensitivity jump magnitude by the normalized denominator to obtain the risk index of the line card slot.

6. The automatic management method for a data center switch system according to claim 1, characterized in that, The target scheduling weight for each line card slot is determined based on the risk index, including: The raw score for the line card slot is calculated based on the risk index; The original scores are normalized to obtain the target scheduling weights for the line card slots.

7. The automatic management method for a data center switch system according to claim 3, characterized in that, Generate change audit hashes based on the initial samples of the line card slots, including: Record the sampling timestamps for the inlet air temperature and the error correction increment rate; Obtain the caller's identity identifier for the initial sample pair; The initial audit hash of the initial sample pair is obtained by hashing the concatenated results of the air inlet temperature, error correction increment rate, sampling timestamp and caller identity of the line card slot. Obtain the configuration change information sent from the control plane to the modular chassis switch; The concatenation result of the initial audit hash, target inflection point temperature, target scheduling weight, and configuration change content is hashed to obtain the change audit hash, so as to achieve secure management of weight configuration adjustment.

8. The automatic management method for a data center switch system according to claim 1, characterized in that, Adjust the port weight configuration of the modular chassis switch based on the target scheduling weight and change audit hash, including: Under the constraints of security management, the updated service scheduling parameters of the port are calculated based on the target scheduling weight; The result of combining the updated service scheduling parameters with the change audit hash is converted into a security configuration command for the modular chassis switch. Under the constraints of security management, the initial service scheduling parameters of the port are configured and updated according to the security configuration command.

Images