Network path switching method and electronic equipment

Abstract

The invention discloses a network path switching method and electronic equipment, and relates to the technical field of data transmission. Network paths of at least two different protocols are configured between a main station and an auxiliary station, in combination with storage performance parameters and network performance parameters of a plurality of network paths, a future performance change trend is predicted by using a preset learning algorithm, and then the weight of each performance parameter is dynamically adjusted. Therefore, comprehensive scoring and optimal path selection of different network paths are realized, a better network path can be predicted and selected in advance before the network performance is reduced or a fault occurs, intelligent rapid switching among different protocol paths is realized, the technical problem of remote replication interruption caused by a single path fault is solved, and the network performance is improved. The technical effects of guaranteeing the continuity and consistency of data transmission and improving the reliability and efficiency of the remote replication network are achieved.

Description

Technical Field

[0001] The present application relates to the technical field of data transmission, and in particular to a network path switching method and electronic device. Background Art

[0002] With the accelerating pace of informatization, data has become a key strategic asset for enterprises, especially critical business data, for which the reliability and security requirements are becoming increasingly stringent. The application of disaster recovery technologies such as remote replication can effectively improve enterprises' risk mitigation capabilities. In the event of emergencies such as natural disasters, hardware failures, or network outages, business systems can be kept running continuously or interruptions can be minimized, thereby achieving better recovery point and recovery time objectives, ensuring business continuity and stability.

[0003] However, existing remote replication technologies suffer from static configurations and single transmission networks. They typically support only a limited number of network protocols and rely on a single path for data transmission between primary and secondary sites. If this path fails, remote replication services are interrupted, posing a high backup risk. When network performance degrades or an outage occurs, it's difficult to automatically and quickly switch to another network protocol. This relies on manual intervention or complex network configuration, making intelligent, dynamic switching impossible. This not only severely impacts remote data replication efficiency and consistency, but also fails to meet the urgent need for efficient and stable remote data replication networks. Summary of the Invention

[0004] The present application provides a network path switching method and electronic device to at least solve the technical problem of remote replication interruption caused by a single path failure in the related art, thereby achieving the technical effect of ensuring the continuity and consistency of data transmission and improving the reliability and efficiency of the remote replication network.

[0005] The present application provides a network path switching method, which is applied to a data transmission system, wherein at least two network paths with different protocols are set between a primary station and a secondary station in the data transmission system, and the network path switching method includes: obtaining storage performance parameters of a storage module in the primary station, and network performance parameters of at least two of the network paths; setting an initial weight for at least one performance parameter, and using a preset learning algorithm to predict a change trend of the performance parameter within a preset time period in the future based on the performance parameter and the corresponding initial weight; adjusting the corresponding initial weight of the performance parameter based on the change trend of the performance parameter to obtain a final weight; calculating a comprehensive score of each of the at least two network paths based on the final weight, the corresponding performance parameter and the change trend; determining a target network path based on the comprehensive score, and switching the data transmission network path between the primary station and the secondary station to the target network path.

[0006] The present application also provides a network path switching device, which is applied to a data transmission system, wherein at least two network paths with different protocols are set between a master station and a slave station in the data transmission system, and the network path switching device includes: a parameter acquisition unit, used to obtain storage performance parameters of a storage module in the master station, and network performance parameters of at least two of the network paths; a prediction unit, used to set an initial weight for at least one performance parameter, and use a preset learning algorithm to predict the change trend of the performance parameter within a preset time period in the future based on the performance parameter and the corresponding initial weight; a weight adjustment unit, used to adjust the initial weight of the corresponding performance parameter according to the change trend of the performance parameter to obtain a final weight; a scoring unit, used to calculate a comprehensive score of at least two of the network paths according to the final weight, the corresponding performance parameter and the change trend; determine a target network path based on the comprehensive score, and switch the data transmission network path between the master station and the slave station to the target network path.

[0007] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned network path switching methods when executing the computer program.

[0008] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned network path switching methods are implemented.

[0009] The present application also provides a computer program product, including a computer program, which implements the steps of any of the above-mentioned network path switching methods when the computer program is executed by a processor.

[0010] Through this application, at least two network paths with different protocols are configured between the primary station and the secondary station, and the storage performance parameters are combined with the network performance parameters of multiple network paths. The preset learning algorithm is used to predict the future performance change trend, and then the weight of each performance parameter is dynamically adjusted to achieve comprehensive scoring and optimal path selection for different network paths. It can predict and select a better network path in advance before network performance degrades or failures occur. Intelligent and rapid switching between different protocol paths solves the technical problem of remote replication interruption caused by single path failure, and achieves the technical effect of ensuring the continuity and consistency of data transmission and improving the reliability and efficiency of the remote replication network. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0012] Figure 1 A flowchart of a network path switching method provided in an embodiment of the present application.

[0013] Figure 2 A flowchart of another network path switching method provided in an embodiment of the present application.

[0014] Figure 3 A schematic diagram of an electronic device provided in an embodiment of the present application.

[0015] Figure 4 A schematic diagram of a computer-readable storage medium provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0017] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.

[0018] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0019] The specific application environment architecture or hardware architecture upon which the execution of the network path switching method depends is described herein. The network path switching method of the present application is applicable to a data transmission system having a primary station and secondary stations. In the system, at least two network paths using different protocols, such as an FC (Fibre Channel) network path and an IP (Internet Protocol) network path, are pre-deployed between the primary station and the secondary stations, connecting the storage modules and transmission equipment of the primary and secondary stations, respectively. The primary station includes a storage module for storing service data, a performance monitoring unit for collecting storage performance parameters and network performance parameters, and a processor for executing the switching method. The secondary stations include a storage module and a data receiving unit corresponding to the primary station. Each network path is composed of a corresponding network interface module, switching equipment, and links. The network paths using different protocols operate independently and can simultaneously establish data transmission connections with the storage modules of the primary and secondary stations. The processor in the primary station uses a collection module to obtain the storage performance parameters of the storage module and the network performance parameters of each network path. Using a preset learning algorithm, it calculates a comprehensive score and controls the switching unit to achieve intelligent and dynamic switching of the data transmission network paths between the primary and secondary stations.

[0020] like Figure 1 The present invention provides a method for switching a network path. The method is described in detail in conjunction with the execution flow of the method. The method is applied to a data transmission system in which at least two network paths with different protocols are set between a primary station and a secondary station. The method includes: S11: Acquire storage performance parameters of a storage module in the master station and network performance parameters of at least two network paths.

[0021] Specifically, it is first necessary to obtain the storage performance parameters of the storage module in the master station. These parameters can reflect the performance status of the master station in the current and recent data processing and transmission processes. Storage performance parameters may include but are not limited to the current read and write bandwidth of the storage module, the number of input and output operations, the average read and write latency, the cache hit rate, and the queue depth. These indicators directly affect the ability to generate, read, and transmit data on the master station side, and are therefore important basic data for evaluating network path switching decisions. The system collects these parameters in real time or periodically through the performance monitoring unit deployed in the master station storage module or the performance statistics interface of the storage module itself, ensuring that the acquired data can accurately reflect the current operating status and load level of the storage module.

[0022] At the same time, it is necessary to obtain the network performance parameters of at least two network paths to reflect the actual performance of different network paths during data transmission. Network performance parameters can include link bandwidth, transmission delay, packet loss rate, jitter, link utilization, etc. This data can be collected through the statistical functions of the network interface module, link detection messages, or network monitoring equipment. Because the multiple network paths deployed between the primary and secondary sites may use different protocols and physical link structures (such as FC networks and IP networks), their performance characteristics and influencing factors vary. Therefore, it is necessary to independently collect and record the performance parameters of each network path. By simultaneously understanding the storage performance of the primary site storage module and the network performance of each network path, a complete input data foundation is provided for subsequent path selection, allowing for a more accurate assessment of the adaptability and advantages and disadvantages of different paths under the current business load.

[0023] In an exemplary embodiment, the storage performance parameters include at least one of the number of input and output operations per second, bandwidth, latency of the storage module of the primary site, and the copy rate progress of remote replication between the primary site and the secondary site; when the network path is a Fibre Channel protocol network path, the network performance parameters include at least one of the Fibre Channel signal-to-noise ratio, cyclic redundancy check error amount, optical module rate, optical module transmit and receive optical power, and bit error rate; when the network path is an Internet Protocol network path, the network performance parameters include network delay, jitter, packet loss rate, and Internet bandwidth utilization during remote replication between the primary site and the secondary site.

[0024] In this embodiment, the selection of storage performance parameters is intended to fully reflect the storage processing capabilities and data output efficiency of the master station when performing remote replication tasks. For example, the number of input and output operations per second can reflect the number of data read and write requests completed by the storage module per unit time, and is an important indicator for measuring the processing capacity of the storage system; bandwidth represents the amount of data that can be transmitted per unit time, which directly determines the upper limit of data output to the network; latency is used to measure the time required from initiating a storage request to completing a response. A lower latency can improve the real-time performance of data transmission; the copy rate progress directly represents the actual progress of the remote replication task between the master station and the slave station, and can comprehensively reflect the impact of both storage and network aspects. These indicators together constitute a multi-dimensional characterization of the operating status of the master station storage module, providing an accurate basis for judging its matching degree with the network path.

[0025] When the FC protocol is used for the network path, performance parameters are primarily selected based on the physical and link layer characteristics of the optical fiber link. The fiber channel signal-to-noise ratio (SNR) reflects the signal quality and interference resistance during transmission. A low SNR may increase the data transmission error rate. The cyclic redundancy check (CRC) error rate reflects the frequency of bit errors during data transmission and is a key indicator of link reliability. The optical module rate limits the theoretical maximum transmission rate of the path. The optical module's transmit and receive power can be used to determine optical link loss and its operating status. The bit error rate directly indicates the link's accuracy during data transmission. These parameters accurately reflect the health of the FC network path at both the physical link and transmission levels.

[0026] When the network path is the IP protocol, the selection of network performance parameters focuses more on the transmission characteristics of the link and the utilization of the service layer. Network latency represents the round-trip time of a data packet between the primary and secondary stations and is an important indicator for measuring real-time performance. Jitter is used to evaluate the fluctuation of packet latency. High jitter can affect the stability of continuous transmission tasks. Packet loss rate directly reflects the proportion of data packets lost during transmission, which can affect data consistency and retransmission overhead. Internet bandwidth utilization reflects the actual use of available bandwidth resources of the IP network in remote replication tasks and can be used to determine whether there is congestion or bandwidth waste on the link. These indicators can provide the system with a basis for evaluating the operating status of IP network paths.

[0027] In an exemplary embodiment, after obtaining the storage performance parameters of the storage module in the master station and the network performance parameters of at least two network paths, it also includes: for at least one performance parameter, using a historical performance parameter set as a training set to calculate the mean, standard deviation, maximum value and minimum value of the performance parameter; the historical performance parameter set is a set of performance parameters selected according to a preset ratio within a preset time period before the current time; and the performance parameters are normalized using the mean, standard deviation, maximum value and minimum value.

[0028] In this embodiment, after obtaining the storage performance parameters of the master station storage module and the network performance parameters of at least two network paths, a statistical analysis process using a historical performance parameter set is performed on at least one performance parameter. The historical performance parameter set is constructed by using the current time as a reference, looking back to a preset time period, and selecting a portion of the performance parameter data collected during that time period according to a preset ratio (e.g., 70%) to form a sample set for analysis. This approach avoids relying solely on instantaneous data at a specific moment, thereby obtaining more representative and trend-based performance information. It also reduces the interference of outliers or short-term fluctuations on subsequent analysis.

[0029] After obtaining a set of historical performance parameters, the mean, standard deviation, maximum, and minimum values ​​of the data in the set are calculated. The mean reflects the overall level of the performance parameter over the selected time period, the standard deviation describes the magnitude of data fluctuation, and the maximum and minimum values ​​provide the extreme range of the parameter within the time period. These statistics not only help identify the stability and volatility of network path or storage module performance but also provide essential reference data for subsequent normalization processing.

[0030] The purpose of normalizing performance parameters using the aforementioned mean, standard deviation, maximum, and minimum values ​​is to map performance parameters of varying dimensions and ranges to a unified numerical range, typically between 0 and 1. This allows for fair comparison and calculation of various performance parameters during subsequent comprehensive path scoring or prediction. This normalization process effectively eliminates the impact of differences in measurement units or magnitudes across performance indicators, making all parameters comparable and consistent in subsequent analysis, providing more accurate and stable basic data for path switching decisions.

[0031] S12: Setting an initial weight for at least one performance parameter, and using a preset learning algorithm to predict a change trend of the performance parameter within a future preset time period based on the performance parameter and the corresponding initial weight.

[0032] Specifically, an initial weight is set for at least one performance parameter to reflect its relative importance in network path selection. The initial weight can be based on empirical values, historical statistical results, or pre-defined policies. For example, in certain business scenarios sensitive to low latency, the initial weight of latency-related performance parameters may be set higher, while in scenarios with high throughput requirements, the initial weight of bandwidth-related parameters may dominate. By properly setting the initial weight, it can be ensured that key performance parameters have a more significant impact on the results in subsequent path evaluations, making the overall score more aligned with actual business needs.

[0033] After obtaining the performance parameters and their initial weights, a pre-set learning algorithm is invoked to leverage current and historical performance data to predict performance parameter trends over a preset time period. This learning algorithm can be based on time series analysis, regression modeling, or other predictive models, specifically LSTM (Long Short-Term Memory) networks. By analyzing parameter fluctuations over historical time periods, the algorithm can infer potential future fluctuations, such as potential increases in latency or decreases in bandwidth. This allows path switching decisions to take into account not only current performance levels but also forward-looking predictions about future performance.

[0034] This approach, combining initial weights with predicted trends, enables the system to proactively mitigate potential risks when selecting network paths, rather than passively responding to current performance conditions. If the prediction indicates a performance parameter is expected to deteriorate in the future, even if current performance remains acceptable, the corresponding network path will have its weight reduced in subsequent calculations due to this trend, thereby improving the foresight of path switching and the reliability of decision-making.

[0035] S13: Adjusting the initial weights of the corresponding performance parameters according to the changing trends of the performance parameters to obtain final weights.

[0036] Specifically, the initial weights set in the previous step are dynamically adjusted based on the changing trends of performance parameters, resulting in final weights that more accurately reflect the impact of current and future performance. Trends reflect the likely trajectory of performance parameters over a predetermined time period, such as increasing latency, decreasing bandwidth, or stabilizing bit error rates. By incorporating this trend information into the weight adjustment process, weight assignments are more closely aligned with the actual operating status of the network and storage, rather than relying solely on statically set initial values.

[0037] The essence of weight adjustment is to increase the weight of a performance parameter when the trend is positive, giving it a larger proportion in the subsequent path's overall score; while in the case of a negative trend, the weight is reduced to reduce its adverse impact on the final decision. For example, if the bandwidth of a certain path is predicted to decrease significantly in the future, even if the current bandwidth is high, the contribution of this parameter to the overall score will be appropriately weakened to prevent short-term advantages from overshadowing long-term risks.

[0038] This trend-based dynamic weight adjustment enables a sensitive response to performance changes, ensuring that the final weight not only reflects the importance of current performance but also incorporates the potential impact of future performance changes. This weight calculation mechanism provides more forward-looking and adaptive input data for subsequent path scoring and selection, thereby improving the accuracy and stability of path switching decisions.

[0039] S14: Calculate the comprehensive scores of at least two network paths according to the final weights, corresponding performance parameters, and change trends.

[0040] Specifically, a comprehensive score is calculated for at least two network paths based on the final weights, corresponding performance parameters, and performance trends. This comprehensive score calculation process aims to uniformly quantify the results of performance parameters of different types and dimensions, and combine their relative importance in the current decision-making scenario to form a numerical assessment of the overall quality of each network path. The final weights play a moderating role in this process, ensuring that the impact on the path score matches the importance of each performance parameter in the current and future operating conditions.

[0041] To calculate the comprehensive score, the normalized performance parameter values ​​are first multiplied by the corresponding final weights to obtain a weighted individual score. The weighted scores of multiple performance parameters are then summed or combined according to a preset formula to arrive at a comprehensive score for each network path. Because the score incorporates both current performance values ​​and projected trends, it reflects not only the path's immediate transmission capacity and stability, but also its potential performance over a predetermined time period.

[0042] The essence of this comprehensive scoring method lies in integrating multi-dimensional performance information into a single, comparable value, enabling network paths with different protocols and physical links to be compared under the same evaluation system. This scoring mechanism objectively and quantitatively assesses the overall transmission quality and adaptability of each path, providing accurate data for the subsequent selection of target network paths.

[0043] S15: Determine the target network path according to the comprehensive score, and switch the data transmission network path between the primary station and the secondary station to the target network path.

[0044] Specifically, the optimal target network path is determined based on the comprehensive scores of each network path calculated in the previous step. A network path with a high comprehensive score indicates superior storage and network performance within the current and future preset time periods, providing greater efficiency and stability for data transmission between the primary and secondary sites. Therefore, by comparing the comprehensive scores of each path, it is possible to objectively identify the transmission path that best suits the current business load and network conditions.

[0045] After determining the target network path, data transmission between the primary and secondary sites is switched from the original network path to the target network path. During the switchover process, the continuity and consistency of data transmission are ensured. By coordinating the storage modules and network interfaces of the primary and secondary sites, a seamless migration of data flow is achieved, avoiding data loss or transmission interruptions during the switchover process.

[0046] The essence of this embodiment lies in using comprehensive scoring results to quantitatively compare multiple network paths. This evaluation directly guides the selection of actual transmission paths, achieving dynamic network path optimization. This approach intelligently selects the optimal path under varying network conditions, improving data transmission reliability and efficiency and ensuring the continuous and stable operation of remote replication services.

[0047] In an exemplary embodiment, the initial weight of the corresponding performance parameter is adjusted according to the changing trend of the performance parameter to obtain the final weight, including: for each performance parameter, judging whether there is a moment when the performance parameter exceeds a first threshold value within a future preset time period according to the changing trend; if there is a moment when the performance parameter exceeds the first threshold value, the initial weight of the performance parameter is increased by a first preset step size to obtain the corresponding final weight.

[0048] In this embodiment, first, based on the changing trends of various performance parameters, possible fluctuations of performance parameters within a preset time period in the future are predicted. By analyzing these trends, it can be determined whether a certain performance parameter is likely to exceed a pre-set first threshold at a certain moment in the future. The first threshold usually represents the key limit allowed for the performance parameter during remote replication or data transmission, such as the maximum tolerance value for delay or the minimum guaranteed value for bandwidth. Exceeding this threshold may affect the stability or efficiency of data transmission. Through this trend analysis, potential performance risk points can be identified in advance, providing a basis for subsequent weight adjustments.

[0049] If the prediction results indicate that a performance parameter will exceed a first threshold within a preset time period in the future, the initial weight of that performance parameter is increased by a first preset step size to obtain the corresponding final weight. This increased weight gives that performance parameter a higher priority in the subsequent network path scoring and selection process, making the path performance associated with that parameter a key factor in decision-making. This proactively favors network paths that demonstrate greater stability or security in key performance parameters, thereby reducing the impact of potential risks on remote replication services.

[0050] The essence of this embodiment is to dynamically associate the future change trends of performance parameters with weights. By weighted processing of performance indicators that may exceed the limit, network path selection can respond to potential problems in advance, realize forward-looking and intelligent path switching, and ensure the continuity and reliability of data transmission between the primary station and the secondary station.

[0051] In an exemplary embodiment, the initial weight of the corresponding performance parameter is adjusted according to the changing trend of the performance parameter to obtain the final weight, including: for each performance parameter, judging according to the changing trend whether there is a moment in the future preset time period when the absolute value of the performance parameter change rate exceeds the second threshold m times; m is a positive integer; if there is a moment when the absolute value of the performance parameter change rate exceeds the second threshold m times, it is judged that the fluctuation of the performance parameter in the future preset time period exceeds expectations, and the initial weight corresponding to the performance parameter is reduced by the second preset step size to obtain the corresponding final weight.

[0052] In this embodiment, based on the changing trend of each performance parameter, the fluctuation of the performance parameter within a preset time period in the future is predicted, and the absolute value of the rate of change is calculated. By determining whether the absolute value of the rate of change of the performance parameter exceeds the second threshold value m times, abnormal fluctuations of the performance parameter that may occur in the future can be identified. Here, m is a positive integer used to set the sensitivity of the judgment, that is, only when the performance parameter changes drastically and frequently enough, is it considered that its fluctuation exceeds expectations, thereby avoiding excessive response to occasional small fluctuations. The second threshold is used to define the acceptable range of the rate of change. When the rate of change exceeds this threshold, it means that the performance parameter may experience large fluctuations in a short period of time, which may affect the stability of data transmission.

[0053] If the prediction results indicate that a performance parameter's absolute rate of change exceeds the second threshold m times within a preset future time period, the parameter is considered to be fluctuating beyond expectations, potentially posing a risk to remote replication or network path stability. To mitigate the impact of this potential risk, the initial weight of the performance parameter is reduced by a second preset step size to obtain a corresponding final weight. This weight reduction reduces the weight of the performance parameter in subsequent comprehensive scoring and network path selection, thereby reducing the possibility of path selection errors caused by excessive fluctuations.

[0054] The essence of this embodiment is to incorporate the future volatility of performance parameters into the weight adjustment mechanism. By identifying potential abnormal changes and appropriately reducing their impact, network path selection can be more robust in the face of unstable performance parameters. This effectively prevents the selection of unstable paths due to short-term fluctuations, and improves the continuity and reliability of data transmission between the primary and secondary stations.

[0055] In an exemplary embodiment, the comprehensive scores of at least two network paths are calculated based on the final weight, the corresponding performance parameters and the change trend, including: Calculate the comprehensive score of each network path;

[0056] Where H is the comprehensive score, i is the serial number of the performance parameter, and n is the total number of storage performance parameters and network performance parameters of the corresponding network path. is the final weight of the i-th performance parameter at time t, is the normalized real-time score of the ith performance parameter at time t, is the maximum score of the i-th performance parameter at time t, is the adjustment coefficient of the predictor, It is the prediction score corresponding to the change trend of the performance parameter at the predicted time t+1.

[0057] In this embodiment, the comprehensive score for each network path is calculated by weighting storage performance parameters and network performance parameters. Specifically, the normalized real-time score of each performance parameter is multiplied by the corresponding final weight and divided by the maximum score for that parameter to obtain a weighted normalized value. This value is then summed across all parameters to reflect the contribution of each performance parameter to the overall performance of the path at the current moment. In this way, performance parameters of different types and dimensions can be comprehensively compared under a unified scoring system, allowing the comprehensive score to objectively reflect the quality of the network path in its current state.

[0058] At the same time, the comprehensive score also introduces the predictive factor , used to adjust the current score based on performance trends at the future time t+1. The LSTM prediction model uses historical and current performance parameter sequences to infer the impact of future performance changes on path performance, ensuring that the comprehensive score reflects not only the current status but also potential future fluctuations. The resulting comprehensive score more comprehensively reflects the performance of network paths in the current and future time periods, providing a reliable basis for selecting target network paths.

[0059] In an exemplary embodiment, before obtaining the storage performance parameters of the storage module in the main station and the network performance parameters of each of at least two network paths, it also includes: setting one of the network paths as the main network path, and setting the other network paths except the main network path as the backup network path; determining the target network path according to the comprehensive score, and switching the data transmission network path between the main station and the secondary station to the target network path, including: when the comprehensive score of the main network path meets the preset requirements, switching the data transmission network path between the main station and the secondary station to the main network path; when the comprehensive score of the main network path does not meet the preset requirements, determining the target network path according to the priority and comprehensive score of each backup network path, and switching the data transmission network path between the main station and the secondary station to the target network path.

[0060] In this embodiment, before obtaining the storage performance parameters of the master station storage module and the network performance parameters of each network path, the available network paths are first divided into roles, with one network path set as the primary network path and the remaining network paths set as backup network paths. The primary network path is typically the preferred path, with higher stability or better historical performance. The backup network path is activated as an alternative path for data transmission when the primary path fails to meet transmission requirements or experiences performance degradation. This primary-backup division allows for quick decision-making during path switching, reducing selection delays and switching risks, while also providing redundancy for the system.

[0061] During network path selection, the primary network path's comprehensive score is calculated and compared with pre-set requirements. If the primary network path's comprehensive score meets the pre-set requirements, indicating that its storage and network performance are capable of supporting business data transmission within the current and future pre-set time periods, the system will prioritize data transmission between the primary and secondary sites over the primary network path to ensure efficient and stable data transmission.

[0062] If the primary network path's overall score fails to meet pre-set requirements, it indicates potential performance risks or inability to meet workload demands. At this point, the backup network paths are evaluated, and the most suitable data transmission path is determined as the target network path based on their overall scores and pre-set priorities. Data transmission between the primary and secondary sites is then switched to this target network path, ensuring the continuity and reliability of remote replication or data transmission tasks.

[0063] The essence of this embodiment lies in the intelligent management of multiple network paths through the division of primary and backup paths and dynamic selection based on comprehensive scoring. Whether the primary path's performance is excellent or fluctuating, the system can quickly determine and switch to the optimal path, ensuring stable and efficient data transmission between the primary and secondary sites, while also improving the overall network's fault tolerance and service continuity.

[0064] In an exemplary embodiment, it also includes: when the comprehensive score of the main network path is lower than the score threshold or when the decline rate of the comprehensive score of the main network path within a preset time interval is higher than the change rate threshold, it is determined that the comprehensive score of the main network path does not meet the preset requirements; when the comprehensive score of the main network path is not lower than the score threshold and the comprehensive score of the main network path does not decline within the preset time interval or the decline rate is not higher than the change rate threshold, it is determined that the comprehensive score of the main network path meets the preset requirements.

[0065] In this embodiment, the primary network path's comprehensive score is first monitored and evaluated in real time to determine whether it can meet data transmission needs within the current and future preset time periods. The comprehensive score reflects the primary network path's overall performance in terms of storage and network performance. Therefore, this evaluation can effectively assess the path's reliability and stability, thus determining whether to switch to a backup network path.

[0066] When the comprehensive score of the primary network path is lower than the preset score threshold, it indicates that the performance of the primary path at the current moment is insufficient to support the business data transmission task, which may lead to increased data transmission delays, reduced throughput, or data consistency risks. At this time, it is determined that the primary network path does not meet the preset requirements, providing a trigger condition for the subsequent selection of a backup path. In addition, even if the current comprehensive score of the primary network path is acceptable, if the rate of decline of its comprehensive score exceeds the rate of change threshold within the preset time interval, it also indicates that the path performance is showing a trend of rapid decline and there is a potential risk. It will also be determined that it does not meet the preset requirements, so protective measures will be taken in advance.

[0067] Conversely, if the primary network path's comprehensive score is at least the threshold and remains stable or its rate of decline is below the threshold over the preset time interval, the primary path's performance is sufficient to meet service requirements both immediately and in the short term. In this case, the primary network path is deemed to meet the preset requirements and is prioritized for continued data transmission, avoiding unnecessary handoffs and improving transmission efficiency and network resource utilization.

[0068] The essence of this embodiment lies in the dual judgment mechanism of comprehensive scoring and thresholds. It not only considers the absolute performance level of the current path, but also conducts dynamic evaluation based on performance change trends. This enables the system to achieve intelligent monitoring and switching decisions for the primary network path while ensuring data transmission stability, thereby improving the foresight and reliability of network path management.

[0069] In an exemplary embodiment, the data transmission network path between the primary station and the secondary station is switched to the target network path, including: when the comprehensive score of the primary network path is lower than the scoring threshold, the data transmission network path between the primary station and the secondary station is switched to the target network path at the current moment when the comprehensive score is determined to be lower than the scoring threshold; when the decline rate of the comprehensive score of the primary network path within a preset time interval is higher than the change rate threshold, and the comprehensive score of the primary network path is not lower than the scoring threshold, at the beginning of the next data transmission cycle, the data transmission network path between the primary station and the secondary station is switched to the target network path.

[0070] In this embodiment, the timing of switching data transmission paths is dynamically determined based on the primary network path's comprehensive score. When the primary network path's comprehensive score falls below a preset threshold, this indicates that the primary path's performance is no longer sufficient to ensure the stability and efficiency of the data transmission task. At this point, data transmission between the primary and secondary stations is immediately switched to the target network path to ensure continuous service data transmission and avoid data loss or transmission interruptions caused by insufficient primary path performance. This instant switching allows for rapid response to performance degradation and mitigates potential risks.

[0071] On the other hand, if the primary network path's comprehensive score remains within the scoring threshold, but its rate of decline exceeds the rate of change threshold within a preset time interval, this indicates a rapid decline in path performance and may not meet future service requirements. To prevent sudden performance issues caused by this downward trend, data transmission is switched to the target network path at the beginning of the next data transmission cycle. This approach prevents potential risks while minimizing frequent switching due to short-term fluctuations, improving system stability.

[0072] The essence of this embodiment lies in combining the absolute value and trend of the comprehensive score to create two corresponding strategies: immediate switching and periodic switching. Immediate switching ensures a rapid response when the primary path performance is significantly insufficient, while periodic switching based on the degradation rate provides proactive management of potential risks, ensuring that data transmission between the primary and secondary stations remains continuous and reliable under all circumstances, while also improving the frequency and efficiency of switching operations.

[0073] In another exemplary embodiment, each time a network path switches, the switch time and specific triggering reason are recorded. Switching reasons may include the primary network path's comprehensive score falling below a preset threshold, the comprehensive score's rate of decline exceeding a threshold, or other performance anomalies. Recording switch times and reasons not only provides data support for subsequent performance analysis but also helps administrators track network status changes and path switching history, improving system management visibility and traceability.

[0074] In another exemplary embodiment, when all available network paths experience failures or abnormal performance, an alert is sent to the administrator, alerting them to the network failure. This ensures that administrators are promptly informed of the abnormality and can take necessary measures. Simultaneously, the performance parameters and status changes of each network path are continuously monitored, allowing immediate action to be taken when the network returns to normal. This real-time monitoring and alerting mechanism ensures data transmission continuity while improving system security and reliability.

[0075] In another exemplary embodiment, once any network path returns to normal, the data replication task is switched back to the restored network path based on the current comprehensive score and path priority. This mechanism dynamically adjusts data transmission paths, achieving efficient utilization of network resources and continuous operation of remote replication services, while minimizing data transmission risks caused by network failures or performance fluctuations.

[0076] like Figure 2In a specific embodiment, the network path switching method based on performance parameters and comprehensive scores is as follows: obtain various performance parameters, normalize the performance parameters, use LSTM dynamic training to determine the final weight, and calculate the comprehensive score of each network path. Then determine whether the main network path is faulty. If it is faulty, determine whether the backup network path is faulty. If the backup path is also faulty, send an alarm; if the main network path is not faulty but the comprehensive score is lower than the backup network path, or the main network path is faulty and the backup network path is not faulty, then switch the data transmission path between the main station and the secondary station to the backup network path with a higher comprehensive score, otherwise maintain the current main network path transmission and end the process. This is another way to implement path switching, which does not conflict with other ways of determining whether a path needs to be switched in the above embodiments.

[0077] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.

[0078] An embodiment of the present application also provides a network path switching device, which is applied to a data transmission system, wherein at least two network paths with different protocols are set between a master station and a slave station in the data transmission system, and the network path switching device includes: a parameter acquisition unit, used to obtain the storage performance parameters of the storage module in the master station, and the network performance parameters of at least two network paths; a prediction unit, used to set an initial weight for at least one performance parameter, and use a preset learning algorithm to predict the change trend of the performance parameter within a preset time period in the future based on the performance parameter and the corresponding initial weight; a weight adjustment unit, used to adjust the initial weight of the corresponding performance parameter according to the change trend of the performance parameter to obtain the final weight; a scoring unit, used to calculate the comprehensive score of at least two network paths respectively based on the final weight, the corresponding performance parameter and the change trend; determine the target network path based on the comprehensive score, and switch the data transmission network path between the master station and the slave station to the target network path.

[0079] In an exemplary embodiment, the weight adjustment unit is specifically used to: for each performance parameter, determine whether there is a moment when the performance parameter exceeds a first threshold value within a future preset time period based on the change trend; if there is a moment when the performance parameter exceeds the first threshold value, increase the initial weight of the performance parameter by a first preset step size to obtain the corresponding final weight.

[0080] In an exemplary embodiment, the weight adjustment unit is specifically used to: for each performance parameter, determine based on the change trend whether there is a moment in the future preset time period when the absolute value of the change rate of the performance parameter exceeds the second threshold m times; m is a positive integer; if there is a moment when the absolute value of the change rate of the performance parameter exceeds the second threshold m times, determine that the fluctuation of the performance parameter in the future preset time period exceeds expectations, and reduce the initial weight corresponding to the performance parameter by a second preset step size to obtain the corresponding final weight.

[0081] In an exemplary embodiment, the scoring unit is specifically configured to: Calculate the comprehensive score of each network path;

[0082] Where H is the comprehensive score, i is the serial number of the performance parameter, and n is the total number of storage performance parameters and network performance parameters of the corresponding network path. is the final weight of the i-th performance parameter at time t, is the normalized real-time score of the ith performance parameter at time t, is the maximum score of the i-th performance parameter at time t, is the adjustment coefficient of the predictor, It is the prediction score corresponding to the change trend of the performance parameter at the predicted time t+1.

[0083] In an exemplary embodiment, it also includes: a primary-backup setting unit, which is used to set one of the network paths as the primary network path and set other network paths except the primary network path as backup network paths; a scoring unit, which is specifically used to switch the data transmission network path between the primary station and the secondary station to the primary network path when the comprehensive score of the primary network path meets the preset requirements; when the comprehensive score of the primary network path does not meet the preset requirements, determine the target network path according to the priority and comprehensive score of each backup network path, and switch the data transmission network path between the primary station and the secondary station to the target network path.

[0084] In an exemplary embodiment, it also includes: a preset requirement determination unit, which is used to determine that the comprehensive score of the main network path does not meet the preset requirements when the comprehensive score of the main network path is lower than the score threshold or when the decline rate of the comprehensive score of the main network path within a preset time interval is higher than the change rate threshold; when the comprehensive score of the main network path is not lower than the score threshold and the comprehensive score of the main network path does not decline within the preset time interval or the decline rate is not higher than the change rate threshold, determine that the comprehensive score of the main network path meets the preset requirements.

[0085] In an exemplary embodiment, the data transmission network path between the primary station and the secondary station is switched to the target network path, including: when the comprehensive score of the primary network path is lower than the scoring threshold, the data transmission network path between the primary station and the secondary station is switched to the target network path at the current moment when the comprehensive score is determined to be lower than the scoring threshold; when the decline rate of the comprehensive score of the primary network path within a preset time interval is higher than the change rate threshold, and the comprehensive score of the primary network path is not lower than the scoring threshold, at the beginning of the next data transmission cycle, the data transmission network path between the primary station and the secondary station is switched to the target network path.

[0086] In an exemplary embodiment, the storage performance parameters include at least one of the number of input and output operations per second, bandwidth, latency of the storage module of the primary site, and the copy rate progress of remote replication between the primary site and the secondary site; when the network path is a Fibre Channel protocol network path, the network performance parameters include at least one of the Fibre Channel signal-to-noise ratio, cyclic redundancy check error amount, optical module rate, optical module transmit and receive optical power, and bit error rate; when the network path is an Internet Protocol network path, the network performance parameters include network delay, jitter, packet loss rate, and Internet bandwidth utilization during remote replication between the primary site and the secondary site.

[0087] In an exemplary embodiment, it also includes: a normalization processing unit, which is used to calculate the mean, standard deviation, maximum value and minimum value of the performance parameter for at least one performance parameter using a historical performance parameter set as a training set; the historical performance parameter set is a set of performance parameters selected according to a preset ratio within a preset time period before the current time; and the performance parameters are normalized using the mean, standard deviation, maximum value and minimum value.

[0088] For the description of the features in the embodiment corresponding to the network path switching device, reference can be made to the relevant description of the embodiment corresponding to the network path switching method, which will not be repeated here.

[0089] like Figure 3 An embodiment of the present application further provides an electronic device, comprising a memory 101 and a processor 102, wherein the memory 101 stores a computer program, and the processor 102 is configured to run the computer program to execute the steps in any of the above-mentioned network path switching method embodiments.

[0090] like Figure 4 An embodiment of the present application further provides a computer-readable storage medium 201, in which a computer program 202 is stored, wherein the computer program 202 is configured to execute the steps of any of the above-mentioned network path switching method embodiments when running.

[0091] In an exemplary embodiment, the computer-readable storage medium 201 may include, but is not limited to, various media that can store the computer program 202, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0092] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above-mentioned network path switching method embodiments are implemented.

[0093] An embodiment of the present application also provides another computer program product, including a non-volatile computer-readable storage medium, the non-volatile computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, implementing the steps in any of the above-mentioned network path switching method embodiments.

[0094] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0095] The above is a detailed introduction to a network path switching method and electronic device provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only intended to help understand the method and core ideas of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A method for switching a network path, characterized in that: Applied to a data transmission system, wherein at least two network paths with different protocols are set between a primary station and a secondary station in the data transmission system, the network path switching method includes: Obtaining storage performance parameters of a storage module in the master station and network performance parameters of at least two of the network paths; Setting an initial weight for at least one performance parameter, and using a preset learning algorithm to predict a change trend of the performance parameter within a preset time period in the future based on the performance parameter and the corresponding initial weight; Adjusting the initial weight of the corresponding performance parameter according to the change trend of the performance parameter to obtain a final weight; Calculating comprehensive scores for at least two of the network paths according to the final weight, the corresponding performance parameter, and the change trend; A target network path is determined according to the comprehensive score, and the data transmission network path between the primary station and the secondary station is switched to the target network path.

2. The network path switching method according to claim 1, characterized in that: Adjusting the initial weight of the corresponding performance parameter according to the change trend of the performance parameter to obtain the final weight includes: For each of the performance parameters, determining, based on the change trend, whether there is a moment in the future preset time period when the performance parameter exceeds a first threshold; If there is a moment when the performance parameter exceeds the first threshold, the initial weight of the performance parameter is increased by a first preset step size to obtain a corresponding final weight.

3. The network path switching method according to claim 1, wherein: Adjusting the initial weight of the corresponding performance parameter according to the change trend of the performance parameter to obtain the final weight includes: For each of the performance parameters, judging based on the change trend whether there are m times in the future preset time period when the absolute value of the change rate of the performance parameter exceeds the second threshold; m is a positive integer; If there are m times when the absolute value of the rate of change of the performance parameter exceeds the second threshold, it is determined that the fluctuation of the performance parameter in the future preset time period exceeds expectations, and the initial weight corresponding to the performance parameter is reduced by the second preset step size to obtain the corresponding final weight.

4. The network path switching method according to claim 1, wherein: Calculating comprehensive scores of at least two network paths based on the final weight, the corresponding performance parameter, and the change trend, including: according to Calculating a comprehensive score for each of the network paths; Wherein, H is the comprehensive score, i is the serial number of the performance parameter, and n is the total number of the stored performance parameters and the network performance parameters of the corresponding network paths. is the final weight of the i-th performance parameter at time t, is the normalized real-time score of the ith performance parameter at time t, is the maximum score of the i-th performance parameter at time t, is the adjustment coefficient of the predictor, It is the prediction score corresponding to the change trend of the performance parameter at the predicted time t+1.

5. The network path switching method according to claim 1, wherein: Before obtaining the storage performance parameters of the storage module in the master station and the network performance parameters of at least two of the network paths, the method further includes: Setting one of the network paths as a primary network path, and setting other network paths except the primary network path as backup network paths; Determining a target network path according to the comprehensive score, and switching the data transmission network path between the primary station and the secondary station to the target network path, including: When the comprehensive score of the primary network path meets a preset requirement, switching the data transmission network path between the primary station and the secondary station to the primary network path; When the comprehensive score of the primary network path does not meet the preset requirement, the target network path is determined according to the priority and comprehensive score of each of the backup network paths, and the data transmission network path between the primary station and the secondary station is switched to the target network path.

6. The network path switching method according to claim 5, characterized in that: Also includes: When the comprehensive score of the primary network path is lower than a score threshold or when the rate of decrease of the comprehensive score of the primary network path within a preset time interval is higher than a change rate threshold, it is determined that the comprehensive score of the primary network path does not meet the preset requirement; When the comprehensive score of the primary network path is not lower than the score threshold and the comprehensive score of the primary network path does not decrease within a preset time interval or the decrease rate is not higher than the change rate threshold, it is determined that the comprehensive score of the primary network path meets the preset requirements.

7. The network path switching method according to claim 6, characterized in that: Switching the data transmission network path between the primary station and the secondary station to the target network path includes: When the comprehensive score of the primary network path is lower than a score threshold, switching the data transmission network path between the primary station and the secondary station to the target network path at the current moment when it is determined that the comprehensive score is lower than the score threshold; When the decline rate of the comprehensive score of the primary network path within a preset time interval is higher than the change rate threshold, and the comprehensive score of the primary network path is not lower than the score threshold, at the beginning of the next data transmission cycle, the data transmission network path between the primary station and the secondary station is switched to the target network path.

8. The network path switching method according to any one of claims 1 to 7, characterized in that: The storage performance parameters include at least one of the number of input and output operations per second, bandwidth, latency of the storage module of the primary site, and the copy rate progress of remote replication between the primary site and the secondary site; When the network path is a Fibre Channel protocol network path, the network performance parameter includes at least one of a Fibre Channel signal-to-noise ratio, a cyclic redundancy check error rate, an optical module rate, an optical module transmit and receive optical power, and a bit error rate; When the network path is an Internet Protocol network path, the network performance parameters include network delay, jitter, packet loss rate, and Internet bandwidth utilization during remote replication between the primary site and the secondary site.

9. The network path switching method according to any one of claims 1 to 7, characterized in that: After obtaining the storage performance parameters of the storage module in the master station and the network performance parameters of at least two of the network paths, the method further includes: For at least one performance parameter, using a set of historical performance parameters as a training set to calculate the mean, standard deviation, maximum, and minimum values ​​of the performance parameter; the set of historical performance parameters is a set of performance parameters selected in a preset ratio within a preset time period before the current time; The performance parameter is normalized using the mean, the standard deviation, the maximum value, and the minimum value.

10. An electronic device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the network path switching method according to any one of claims 1 to 9 when executing the computer program.

Images