A network optimization method and device, a storage medium and a computer program product

Abstract

The application provides a network optimization method and device, a storage medium and a computer program product. The method comprises the following steps: acquiring a bandwidth utilization rate of a network interface of a cloud service instance in a current time period; determining a threshold corresponding to the current time period according to a seasonal variation analysis result of historical bandwidth usage data and a bandwidth usage analysis result; determining a current bandwidth usage state according to the bandwidth utilization rate and the threshold; monitoring packet loss data of the network interface of the cloud service instance; determining network optimization information according to the current bandwidth usage state and the packet loss data; the network optimization information comprises at least one of the following: traffic optimization, network configuration optimization, speed limiting strategy adjustment and bandwidth capacity adjustment.

Description

Technical Field

[0001] This application relates to cloud computing service technology, and more particularly to a network optimization method and apparatus, storage medium, and computer program product. Background Technology

[0002] With the continuous maturation and popularization of cloud computing technology, more and more enterprises and individual users are relying on cloud services. In the era of cloud computing and big data, the stability and reliability of network services are crucial. However, existing network diagnostic methods often lack flexibility and intelligence, and cannot effectively handle complex network problems. Summary of the Invention

[0003] This application provides a network optimization method and apparatus, a storage medium, and a computer program product.

[0004] The technical solution of this application is implemented as follows: In a first aspect, this application proposes a network optimization method, the method comprising: Get the bandwidth utilization rate of the network interface of the cloud service instance in the current time period; The threshold corresponding to the current time period is determined based on the seasonal variation analysis results of historical bandwidth usage data and the bandwidth usage analysis results; and the current bandwidth usage status is determined based on the bandwidth utilization rate and the threshold. Monitor packet loss data on the network interfaces of cloud service instances; Based on the current bandwidth usage status and the packet loss data, network optimization information is determined; the network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment.

[0005] Optionally, determining the threshold corresponding to the current time period based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data includes: Determine the first seasonal cycle to which the current time period belongs; Based on the seasonal variation analysis results, the first bandwidth baseline value corresponding to the first seasonal cycle is determined; Based on the bandwidth usage analysis results, predict the bandwidth usage value for the next time period; The threshold is determined based on the first bandwidth baseline value and the bandwidth usage value.

[0006] Optionally, the method further includes: Abnormal historical bandwidth usage data is identified and removed from the historical bandwidth usage data to obtain normal historical bandwidth usage data. Calculate the historical average bandwidth usage data for each time period based on the normal historical bandwidth usage data; Based on the historical average bandwidth usage data for each time period, identify each seasonal period and determine the corresponding bandwidth baseline value for each seasonal period. By using historical average bandwidth usage data for each time period, a bandwidth usage trend is fitted as the result of the bandwidth usage analysis.

[0007] Optionally, after monitoring packet loss data on the network interface of the cloud service instance; before determining network optimization information based on the current bandwidth usage status and the packet loss data, the method further includes: Obtain the rate limiting policy set on the network interface of the cloud service instance, analyze the rate limiting policy, and generate analysis results; the analysis results include, but are not limited to, at least one of the following: the purpose of setting the rate limiting policy and the impact of implementing the rate limiting policy; Based on the packet loss data and / or the rate limiting strategy, the cause of packet loss is generated. The step of determining network optimization information based on the current bandwidth usage status and the packet loss data includes: Based on the reasons for packet loss, the analysis results, and the current bandwidth usage status, the network optimization information is determined.

[0008] Optionally, the method further includes: By performing packet inspection on the data packets flowing through the cloud service instance, it can be determined whether to switch network paths; Once the network path to be switched is determined, the problem node is identified based on the network topology and network status monitoring information; A shortest path algorithm is used to determine a candidate path from the problem node to the destination node; the destination node is determined based on the destination address identifier carried in the data packet. Determine the path evaluation information of the candidate paths, and select a replacement path from the candidate paths based on the path evaluation information; the path evaluation information is determined by at least one of the following: path length, link performance overhead, link load overhead, and historical failure loss index. Based on the replacement path, generate path switching suggestions and / or path switching guidance information.

[0009] Optionally, the method further includes: The first data is inspected, and an alarm message is generated when an anomaly is detected in the first data. The first data includes, but is not limited to, at least one of the following: cloud service instance status, resource status data of resources bound to the cloud service instance, cloud host status data associated with the cloud service instance, order status data of orders related to the cloud service instance, location data of the IP address of the cloud service instance, and blocking status data of the IP address of the cloud service instance.

[0010] Optionally, the method further includes: Detect the insufficient metering deduction status of orders related to cloud service instances and whether the insufficient metering deduction status has expired; Predicted metering deduction data for orders related to cloud service instances; Based on the insufficient metering deduction status, whether the insufficient metering deduction status has expired, and the expected metering deduction data, a first optimization strategy is generated; wherein, the first optimization strategy includes, but is not limited to, at least one of the following: adjusting the size of the cloud service instance, or disabling unused resources bound to the cloud service instance.

[0011] Secondly, this application proposes a network optimization apparatus, which includes a processor and a memory; the processor executes the running program stored in the memory to implement the above-mentioned network optimization method.

[0012] Thirdly, this application proposes a storage medium on which a computer program is stored, which, when executed by a processor, implements the above-mentioned network optimization method.

[0013] Fourthly, this application proposes a computer program product, including a computer program that implements the above-mentioned network optimization method when executed by a processor.

[0014] The embodiments of this application have the following beneficial effects: Based on the analysis results of seasonal changes in historical bandwidth usage data and bandwidth usage analysis results, the threshold of bandwidth utilization is dynamically adjusted, which improves the accuracy of determining the current bandwidth usage status. Combined with packet loss data monitored on the network interface of cloud service instances, for more complex network problems, the causes of network congestion can be determined more comprehensively and accurately, thereby proposing more accurate network optimization information and improving the flexibility and intelligence of network diagnosis. Attached Figure Description

[0015] Figure 1 A flowchart illustrating a network optimization method provided in this application embodiment; Figure 2 A flowchart illustrating an exemplary bandwidth utilization detection method provided in this application embodiment; Figure 3 A schematic flowchart illustrating an exemplary packet loss detection method provided in this application embodiment; Figure 4 A schematic flowchart illustrating an exemplary bandwidth capacity diagnosis method provided in this application embodiment; Figure 5 A flowchart illustrating an exemplary network connectivity diagnosis method provided in this application embodiment; Figure 6A schematic flowchart illustrating an exemplary basic diagnostic method provided in this application embodiment; Figure 7 A flowchart illustrating an exemplary order diagnosis method provided in this application embodiment; Figure 8 A schematic diagram of an exemplary resilient network diagnostic and optimization system provided in this application embodiment; Figure 9 A schematic diagram of the structure of a network optimization device provided in this application embodiment. Figure 1 ; Figure 10 A schematic diagram of the structure of a network optimization device provided in this application embodiment. Figure 2 .

[0016] It should be noted that the terms "first" and "second" mentioned above are only used to distinguish between different options and do not represent the degree of superiority or inferiority of the options or their priority in the implementation process. Detailed Implementation

[0017] In order to gain a more detailed understanding of the features and technical content of the embodiments of this application, the implementation of the embodiments of this application will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for reference and illustration only and are not intended to limit the embodiments of this application.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] In the following description, references to "some embodiments" refer to a subset of all possible embodiments. It is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. It should also be noted that the terms "first, second, third" used in the embodiments of this application are merely for distinguishing similar objects and do not represent a specific ordering of objects. It is understood that "first, second, third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0020] This application provides a network optimization method, such as... Figure 1 As shown, the method may include: S101. Obtain the bandwidth utilization rate of the network interface of the cloud service instance within the current time period.

[0021] The network optimization method proposed in this application can be applied to cloud service scenarios such as cloud storage performance monitoring, database optimization and diagnosis, mobile device management (MDM) systems, and software as a service (SaaS) services based on cloud platforms.

[0022] In this embodiment of the application, real-time traffic data of the network interface of the cloud service instance is collected, including the number and bytes of inbound data packets and the number and bytes of outbound data packets.

[0023] In this embodiment of the application, the bandwidth utilization rate in the current time period is calculated, as shown in formula (1).

[0024] Bandwidth utilization = (Real-time traffic data within the current time period / Bandwidth capacity) 100% (1) S102. Determine the threshold corresponding to the current time period based on the seasonal variation analysis results of historical bandwidth usage data and the bandwidth usage analysis results; and determine the current bandwidth usage status based on the bandwidth utilization rate and the threshold.

[0025] It should be noted that the thresholds corresponding to the current time period can include low and high thresholds. The initial values ​​of the low and high thresholds can be set based on the normal fluctuation range of bandwidth utilization. For example, if the normal fluctuation range of bandwidth utilization is 20%-80% (determine the acceptable minimum and maximum bandwidth utilization based on the performance requirements of network services. Consider the bandwidth requirements of critical services and ensure that critical services have sufficient bandwidth guarantees), the initial value of the low threshold can be set to 30% as a warning level, and the initial value of the high threshold can be set to 90%, indicating that the bandwidth is close to saturation and action needs to be taken.

[0026] It should be noted that the threshold for the current time period can be adaptively adjusted based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data.

[0027] In this embodiment of the application, the first seasonal cycle to which the current time period belongs is determined; based on the seasonal change analysis results, the first bandwidth baseline value corresponding to the first seasonal cycle is determined; based on the bandwidth usage analysis results, the bandwidth usage value of the next time period is predicted; and based on the first bandwidth baseline value and the bandwidth usage value, a threshold is determined.

[0028] In one embodiment, the current time period is obtained, such as what day of the week the current time period is, and then the first seasonal cycle to which the current time period belongs (such as what day of the week) is determined. The seasonal average bandwidth usage (first bandwidth baseline value) corresponding to the first seasonal cycle is extracted from the seasonal variation analysis results, and the bandwidth usage value of the next time period is predicted based on the bandwidth usage analysis results.

[0029] In one embodiment, the first bandwidth baseline value and the bandwidth usage value are input into formula (2) to calculate the threshold.

[0030] dynamic_threshold=seasonal_avg+sensitivity (predicted_trend-seasonal_avg) (2) Where dynamic_threshold is the threshold, seasonal_avg is the first bandwidth baseline value, predicted_trend is the bandwidth usage value for the next time period, and sensitivity is the sensitivity parameter.

[0031] It should be noted that (predicted_trend-seasonal_avg) represents the trend adjustment amount, used to calculate the difference between the predicted future trend value (predicted_trend) and the current seasonal baseline value (seasonal_avg). This difference quantifies the degree of deviation of the long-term trend from the current typical level (positive values ​​indicate expected growth, and negative values ​​indicate expected decline).

[0032] It should be noted that the sensitivity parameter is a configurable adjustment factor used to control the strength of the dynamic threshold's response to changes in the predicted trend. A higher sensitivity value makes the threshold follow changes in the predicted trend more aggressively (for example, if the predicted trend is significantly higher than the current seasonal average, high sensitivity will generate a higher threshold to prepare for expected growth); a lower sensitivity value makes the threshold more likely to remain near the current seasonal average level.

[0033] It should be noted that the optimal value of the sensitivity parameter can be determined through A / B testing. Specifically, different sensitivity values ​​are applied to different network regions or time periods, and then network performance metrics (such as congestion rate, resource utilization, user experience score, etc.) under these configurations are monitored and compared. The best-performing value is then selected as the standard parameter for the algorithm.

[0034] It should be noted that the seasonal baseline value (seasonal_avg) and the trend adjustment amount (sensitivity) adjusted for sensitivity factors are used to compare the seasonal baseline value (seasonal_avg). The sum of (predicted_trend-seasonal_avg) yields the final dynamic threshold. This threshold incorporates historical periodic patterns (first bandwidth baseline value) and proactively integrates the predicted long-term direction of change (trend adjustment), while providing regulatory flexibility through a sensitivity parameter.

[0035] In this embodiment, abnormal historical bandwidth usage data is identified and removed from historical bandwidth usage data to obtain normal historical bandwidth usage data; the historical average bandwidth usage data for each time period is calculated based on the normal historical bandwidth usage data; based on the historical average bandwidth usage data for each time period, each seasonal period is identified and the corresponding bandwidth baseline value for each seasonal period is determined; and the bandwidth usage trend is fitted using the historical average bandwidth usage data for each time period as the bandwidth usage analysis result.

[0036] In one embodiment, bandwidth usage data is collected at fixed time intervals (e.g., hourly or daily). The collected data is stored in a database as historical bandwidth usage data for long-term analysis and trend identification. The collected historical bandwidth usage data includes, but is not limited to, information such as total traffic, peak traffic, and average traffic.

[0037] In one embodiment, abnormal historical bandwidth usage data (such as extreme values ​​caused by sudden failures or special events) are identified and excluded from historical bandwidth usage data to ensure that these abnormal historical bandwidth usage data do not interfere with the identification of seasonal cycles and long-term trends, as well as the final threshold calculation.

[0038] In one embodiment, normal historical bandwidth usage data is processed by grouping it according to a preset time period and calculating the historical average bandwidth usage data for each time period.

[0039] In one embodiment, the preset time period can be a date; correspondingly, the historical average bandwidth usage data for each time period can be the daily average bandwidth usage.

[0040] In one embodiment, based on historical average bandwidth usage data for each time period, seasonal cycles are identified and corresponding bandwidth baseline values ​​are determined. For example, the daily average bandwidth usage is grouped by day of the week (Monday to Sunday) and the average value of each group is calculated, thereby obtaining a seasonal baseline value (seasonal_avg) (i.e., bandwidth baseline value) reflecting the weekly cycle pattern (seasonal cycle). This effectively captures the traffic variation patterns between weekdays and weekends, specific holidays, or seasonalities.

[0041] In one embodiment, to identify long-term trends, historical average bandwidth usage data for each time period is used to fit a bandwidth usage trend. For example, the time series (such as daily average usage arranged in chronological order) is converted into a numerical indexed sequence (e.g., 0 represents day one, 1 represents day two, and so on). Statistical methods such as linear regression are used to model this sequence, fitting a trend line of bandwidth usage over time (bandwidth usage trend). Based on this model, the bandwidth usage value for the next time period (the following day) can be predicted (predicted_trend).

[0042] In this embodiment of the application, the bandwidth utilization rate within the current time period is continuously recorded and compared with the threshold corresponding to the current time period.

[0043] In one embodiment, if the bandwidth utilization is between a low threshold and a high threshold, it indicates that the network is in a normal state and is operating normally, requiring no special attention.

[0044] In one embodiment, if bandwidth utilization approaches or reaches a low threshold, it indicates a warning state, suggesting a potential network problem requiring further monitoring. The cause may need to be investigated, such as whether some non-critical services are consuming excessive bandwidth.

[0045] In one embodiment, if bandwidth utilization reaches or exceeds a high threshold, it indicates an emergency state, suggesting that the network may be overloaded and immediate action is required (characterizing a network optimization message).

[0046] In one embodiment, the following actions may be performed: Network configuration optimization, such as adjusting Quality of Service (QoS) settings.

[0047] Adjusting bandwidth capacity, such as increasing bandwidth capacity, upgrading network interfaces, or adding additional network links.

[0048] Notify the network administrator to conduct further troubleshooting and performance optimization.

[0049] It should be noted that if a temporary peak is detected, you can consider temporarily increasing bandwidth or implementing traffic shaping strategies, such as temporarily increasing traffic.

[0050] Based on the above embodiments, this application proposes a bandwidth utilization detection method, see [link to relevant documentation]. Figure 2 ,include: 1. Collect real-time traffic data from network interfaces.

[0051] 2. Calculate bandwidth utilization.

[0052] 3. Compare bandwidth utilization with the threshold.

[0053] 4. Dynamically determine the threshold.

[0054] 5. Display status, including normal status, warning status, and emergency status.

[0055] 6. Implementation measures.

[0056] The tests include optimizing network configuration, increasing bandwidth capacity, and notifying network administrators.

[0057] S103, Monitor packet loss data on the network interface of the cloud service instance.

[0058] In one embodiment, the packet loss data may include the number of lost packets, type, timestamp, duration, etc., wherein the type includes: TCP packet loss, UDP packet loss, etc.

[0059] Furthermore, after monitoring packet loss data on the network interface of the cloud service instance, the rate limiting policy set on the network interface of the cloud service instance can be obtained, and the rate limiting policy can be analyzed to generate analysis results. The analysis results include, but are not limited to, at least one of the following: the purpose of setting the rate limiting policy, the impact of implementing the rate limiting policy; and the analysis based on packet loss data and / or rate limiting policy to generate the cause of packet loss.

[0060] In one embodiment, the acquired rate limiting policy may include a rate limiting value, application time, triggering conditions, etc.

[0061] In one embodiment, the purpose of setting a rate limiting policy may include controlling costs, ensuring bandwidth for critical services, or other purposes.

[0062] In one embodiment, the impact of implementing the rate limiting policy includes the impact of changes in bandwidth utilization after the implementation of the rate limiting policy, the impact of changes in traffic patterns, etc.

[0063] In one embodiment, the impact of changes in bandwidth utilization after the implementation of the rate limiting policy can be whether the rate limiting has caused the bandwidth utilization to approach or reach the rate limiting value.

[0064] In one embodiment, the impact of a change in traffic patterns could be whether a specific type of traffic (such as video streaming or large file transfers) is affected by rate limiting.

[0065] In one embodiment, it can be checked whether the time of the packet loss event coincides with the implementation time of the rate limiting policy to determine whether the rate limiting caused network congestion, thereby causing packet loss (the cause of packet loss). If the rate limit value is set too low, it may lead to network congestion and packet loss.

[0066] In one embodiment, the response of different protocols (Transmission Control Protocol (TCP), User Datagram Protocol (UDP), etc.) to rate limiting can be analyzed. For example, TCP will attempt to retransmit when packets are lost, while UDP will not, which may affect packet loss performance.

[0067] In one embodiment, the network device's buffer size can be checked to ensure it is sufficient to handle traffic fluctuations after rate limiting.

[0068] S104. Determine network optimization information based on the current bandwidth usage status and packet loss data; network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment.

[0069] In this embodiment of the application, network optimization information can be determined based on the cause of packet loss, analysis results, and current bandwidth usage status.

[0070] In this embodiment, network optimization information can be generated based on the cause of network congestion. The specific selection can be made according to the actual situation, and this embodiment does not impose any specific limitations.

[0071] In one embodiment, network congestion may include packet loss, high bandwidth utilization, etc. The specific cause can be selected according to the actual situation, and this application embodiment does not make specific limitations.

[0072] In one embodiment, if the analysis shows that packet loss is not directly related to the rate limiting policy, it may be necessary to further investigate other factors that could cause packet loss. If packet loss is directly related to the rate limiting policy, it may be necessary to adjust the rate limiting policy to reduce packet loss. Adjusting the rate limiting policy may include: increasing the rate limit value, optimizing the rate limiting time window, or changing the rate limiting conditions.

[0073] In one embodiment, if the network congestion is caused by high bandwidth utilization, network configuration optimization can also be performed, such as increasing the buffer size or adjusting the queue management algorithm.

[0074] Based on the above embodiments, this application proposes a packet loss detection method, see [link to relevant documentation]. Figure 3 ,include: 1. Collect traffic data on network interfaces.

[0075] 2. Check if any restriction policies are set on the network interface.

[0076] 3. If a rate limiting policy is detected on the network interface, analyze the relationship between the rate limiting setting and packet loss.

[0077] Specifically, it can perform processes such as data collection, analysis of rate limiting strategies, correlation analysis, and packet loss cause analysis.

[0078] Data collection includes: packet loss data, network traffic data, and restriction policy data.

[0079] Analyzing rate limiting policies includes: 1. Policy review: Reviewing the settings of the rate limiting policy to understand whether its purpose is to control costs, ensure bandwidth for critical services, or for other reasons. 2. Policy effectiveness: Evaluating the effectiveness of the rate limiting policy after its implementation, including changes in bandwidth utilization and traffic patterns.

[0080] The correlation analysis includes: 1. Time correlation: checking whether the time of packet loss events coincides with the implementation time of the rate limiting policy. 2. Bandwidth utilization analysis: analyzing the changes in bandwidth utilization before and after the implementation of the rate limiting policy to determine whether the rate limiting caused the bandwidth utilization to approach or reach the rate limit value. 3. Traffic pattern comparison: comparing the traffic patterns before and after the rate limiting to see if specific types of traffic (such as video streaming or large file transfers) were affected by the rate limiting.

[0081] Packet loss analysis includes: 1. Bandwidth limiting impact: Determine if rate limiting caused network congestion, thus leading to packet loss. Setting the rate limit too low may cause network congestion and packet loss. 2. Protocol characteristics: Analyze the response of different protocols (TCP, UDP, etc.) to rate limiting. For example, TCP will attempt to retransmit lost packets, while UDP will not, which may affect packet loss. 3. Buffer size: Check if the network device's buffer size is sufficient to cope with traffic fluctuations after rate limiting.

[0082] Understandably, by dynamically adjusting the bandwidth utilization threshold based on the seasonal variation analysis results of historical bandwidth usage data and bandwidth usage analysis results, the accuracy of determining the current bandwidth usage status is improved. Combined with packet loss data monitored on the network interfaces of cloud service instances, more comprehensive and accurate determination of the causes of network congestion can be made for more complex network problems, thereby providing more accurate network optimization information and improving the flexibility and intelligence of network diagnosis.

[0083] Based on the above embodiments, this application proposes a bandwidth capacity diagnosis method, see [link to relevant documentation]. Figure 4 ,include: 1. Detect the bandwidth utilization of inbound and outbound traffic.

[0084] It should be noted that this is used for a preliminary assessment of the network's operational status.

[0085] 2. Adaptive threshold adjustment.

[0086] It should be noted that this can ensure the threshold is always optimal, reducing false alarms and false negatives.

[0087] 3. Bandwidth limiting and packet loss detection.

[0088] It should be noted that in-depth analysis of detected abnormal states, such as bandwidth utilization and packet loss, is required.

[0089] 4. Provide traffic guidance.

[0090] It should be noted that traffic routing provides network administrators with specific suggestions for optimizing network performance, which may include recommended traffic optimization measures and suggestions for improving network performance.

[0091] Furthermore, embodiments of this application can also implement a network connectivity diagnosis method, see details below. Figure 5 ,include: 1. By performing packet inspection on data packets flowing through cloud service instances, determine whether to switch network paths.

[0092] In one embodiment, to address issues such as inaccessibility and packet loss when accessing cloud service instances, the system checks the instance status, bound resource status, basic Distributed Denial of Service (DDoS) protection, security group rules, network access control list (ACL) configuration, various statuses, and self-check items.

[0093] It should be noted that this application proposes a network problem diagnosis scheme that integrates deep packet inspection (DPI) and intelligent routing analysis. Based on DPI technology, combined with application layer protocol identification and real-time network status monitoring, it achieves accurate identification of network problems and intelligent routing decisions by comprehensively analyzing data packet content and link performance indicators. This enables precise diagnosis of network problems and intelligent route selection. This application not only deeply identifies specific application layer problems but also dynamically evaluates and automatically suggests optimal alternative routes, significantly improving the automation level, response speed, and problem-solving efficiency of network management.

[0094] In one embodiment, network packet capture tools (such as Wireshark or tcpdump) can be used to capture data packets flowing through the cloud service instance.

[0095] In one embodiment, the packet header information can be analyzed to confirm the protocol type, source / destination Internet Protocol (IP) address, port number, etc. The packet content can be examined to identify specific application layer protocols (such as Hypertext Transfer Protocol (HTTP), Hypertext Transfer Protocol Secure (HTTPS), File Transfer Protocol (FTP), etc.).

[0096] In one embodiment, patterns of packet loss can be identified, such as continuous packet loss or specific types of packet loss. The packet content is analyzed to determine if there are application-layer issues, such as incorrect request formats or response timeouts.

[0097] In one embodiment, when deep packet inspection or other monitoring methods confirm the existence of network path-related performance issues (such as high latency, high packet loss rate) or link / node failures, and a better path needs to be found, a network path switching is determined, triggering an intelligent routing analysis process.

[0098] 2. Once the network path to be switched is determined, identify the problem node based on the network topology and network status monitoring information.

[0099] In one embodiment, the complete topology information of the current network can be obtained in real time through the network management application programming interface (API) provided by the cloud service. This includes all network nodes (such as routers, switches, and gateways) and links (physical lines, virtual tunnels, virtual private cloud (VPC) peering connections, etc.) and the connections between them.

[0100] In one embodiment, real-time performance metrics (network status monitoring information) of each link in the network can be continuously monitored, including but not limited to: Latency; Packet Loss Rate; Bandwidth utilization; Link availability status (Up / Down), etc.

[0101] In one embodiment, the specific faulty node causing the problem (e.g., a router failure or a physical link outage) is identified based on network topology and network status monitoring information. The scope of the impact of the fault on overall network connectivity and performance is then assessed.

[0102] 3. Use the shortest path algorithm to determine the candidate path from the problem node to the destination node; the destination node is determined based on the destination address identifier carried in the data packet.

[0103] In one embodiment, an enhanced shortest path algorithm (based on Dijkstra's principle) can be used to compute all possible alternative candidate paths from the problem node (the network entry point of the problem instance) to the target node (the access destination).

[0104] 4. Determine the path evaluation information of the candidate paths, and select a replacement path from the candidate paths based on the path evaluation information; the path evaluation information is determined by at least one of the following: path length, link performance overhead, link load overhead, and historical failure loss indicators.

[0105] In one embodiment, path evaluation information is calculated for each candidate path. The path evaluation information is calculated by dynamically summing the following factors across all links on the path: Path length (number of hops): Number of nodes (base weight).

[0106] Link performance overhead: Calculated based on link performance data such as real-time / recent latency, packet loss rate, and bandwidth (the worse the performance, the greater the overhead).

[0107] Link load overhead: Calculated based on the real-time bandwidth utilization of the link (the higher the load, the greater the overhead, which is adjusted by the load cost factor (LOAD_COST_FACTOR)).

[0108] Historical fault loss metric: Calculated based on the historical fault frequency / stability of the link (the more fault records, the greater the overhead, which is adjusted by the fault cost factor (FAULT_COST_FACTOR)).

[0109] In one embodiment, the path evaluation information of all candidate paths is compared, and the candidate path with the lowest path evaluation information is selected as the best replacement path.

[0110] 5. Generate path switching suggestions and / or path switching guidance information based on the replacement path.

[0111] In one embodiment, if the alternative path is significantly better than the current problematic path (e.g., much lower cost or meeting performance thresholds), a path switching suggestion containing detailed alternative path information is automatically generated; and / or, path switching guidance information is provided to support automatic switching (executed by the system) or manual switching (executed with administrator approval).

[0112] It should be noted that if there are multiple low-cost candidate paths, one can be selected as the replacement path based on the load balancing strategy.

[0113] It is understood that this application proposes an intelligent routing analysis scheme based on deep packet inspection, which combines application layer protocol identification with real-time network status monitoring. By comprehensively analyzing packet content and link performance, it can more comprehensively identify and resolve network problems, including packet content analysis and possible alternative route suggestions, thus achieving accurate diagnosis and intelligent routing selection for network problems. Real-time monitoring and fault diagnosis: It emphasizes real-time monitoring of network link performance indicators and uses the shortest path algorithm combined with network status assessment for fault diagnosis, ensuring that network problems can be quickly and accurately located and resolved.

[0114] Furthermore, embodiments of this application can also implement basic diagnostic methods, as detailed in the following examples. Figure 6 ,include: 1. Detect the first data and generate an alarm message when an anomaly is detected in the first data; the first data includes, but is not limited to, at least one of the following: cloud service instance status, resource status data of resources bound to the cloud service instance, cloud host status data associated with the cloud service instance, order status data of orders related to the cloud service instance, location data of the IP address of the cloud service instance, and blocking status data of the IP address of the cloud service instance.

[0115] It should be noted that basic diagnostics are mainly used to detect the basic status of cloud service instances and ensure that cloud service instances are running normally.

[0116] In one embodiment, instance status detection is used to check whether a cloud service instance is running normally. Specifically, this includes: encapsulating an interface to query the instance's running status; comparing the instance status with the expected status (e.g., running, stopped); and if the status does not match, recording the problem and generating a warning.

[0117] In one embodiment, the bound resource status detection is used to confirm whether the resources (such as disks and networks) bound to the instance are functioning properly. Specifically, this includes: querying all resources bound to the instance based on the instance identification (ID); checking the running status of each resource in turn; and for abnormal resources, recording detailed information and generating a warning.

[0118] In one embodiment, the associated cloud host status detection is used to check the status of cloud hosts if an instance is associated with it. Specifically, this includes: querying all cloud hosts associated with the instance based on the instance ID; checking the running status and performance metrics of each cloud host; and for hosts with abnormal status, logging the problem and generating a warning.

[0119] In one embodiment, order status detection is used to verify the order status related to the instance to ensure that the service has not expired. Specifically, this includes: retrieving instance-related order information based on the instance ID; checking the order status (e.g., valid, expired, frozen, etc.); and if the order status is abnormal, recording the issue and notifying the user.

[0120] In one embodiment, IP geolocation detection is used to confirm the geolocation of the IP address used by the instance in order to troubleshoot geolocation-related issues. Specifically, this includes: obtaining the IP address used by the instance; querying the geolocation information of the IP address; comparing the IP geolocation with the expected one, and recording the problem if there is a discrepancy.

[0121] In one embodiment, IP blocking status detection is used to check whether the IP address of an instance is blocked. Specifically, this includes: using a third-party service or internal system to check the blocking status of the IP address; if the IP is blocked, recording detailed information and generating a warning.

[0122] Furthermore, embodiments of this application can also implement an order diagnosis method, as detailed in [link to specific examples]. Figure 7 ,include: 1. Detect the insufficient metering deduction status of cloud service instance related orders and whether the insufficient metering deduction status has expired.

[0123] In one embodiment, detecting the insufficient metering deduction status of orders related to cloud service instances and whether the insufficient metering deduction status has expired only supports internet accounts.

[0124] In one embodiment, order diagnostics are primarily used to monitor the financial status of cloud service instances to ensure that the continuous operation of service instances is not interrupted due to cost issues.

[0125] In one embodiment, the detection of insufficient metering deduction for cloud service instance related orders includes: querying the billing API based on the instance ID to obtain the current financial status of the instance; comparing the instance's order status with the billing rules to determine whether the instance is in arrears (insufficient metering deduction); if the instance is in arrears, recording the arrears amount, the start time of the arrears, and the reason for the arrears.

[0126] In one embodiment, the detection of whether the metering deduction is due includes: extracting relevant information of the overdue payment notification from the billing API, including the overdue payment due date; calculating the time difference between the current time and the overdue payment due date to determine the remaining processing time; and generating a warning notification if the overdue payment due date is approaching or has expired, and notifying the relevant administrators or users.

[0127] 2. Expected metering deduction data for orders related to cloud service instances.

[0128] In one embodiment, predictive analytics is used to forecast future metering deductions to help plan budgets and resources, such as reducing resource usage.

[0129] 3. Based on the insufficient metering deduction status, whether the insufficient metering deduction status has expired, and the expected metering deduction data, generate a first optimization strategy; wherein, the first optimization strategy includes, but is not limited to, at least one of the following: adjusting the size of the cloud service instance, or closing the unused resources bound to the cloud service instance.

[0130] It is understood that this application embodiment designs an intelligent cloud service self-service troubleshooting system, including multiple modules such as basic diagnosis, bandwidth capacity diagnosis, cost diagnosis, and network connectivity diagnosis, which can comprehensively diagnose and intelligently resolve network problems. It integrates intelligent diagnosis, automatic optimization, user behavior analysis, and a visual interface, providing the ability to quickly locate and intelligently resolve problems that may occur during cloud resource usage. It can quickly and accurately identify and resolve problems in cloud services, thereby improving service stability and reliability, and enhancing user satisfaction.

[0131] Based on the above embodiments, this application proposes a resilient network diagnosis and optimization system, see [link to relevant documentation]. Figure 8The system includes: a user interface, a basic diagnostic module, a bandwidth capacity diagnostic module, a cost diagnostic module, a network connectivity diagnostic module, data storage and processing, and system management and configuration. The user interface supports information display and collection; the basic diagnostic module implements basic diagnostic methods; the bandwidth capacity diagnostic module implements bandwidth capacity diagnostic methods; the network connectivity diagnostic module implements network connectivity diagnostic methods; and the cost diagnostic module implements order diagnostic methods. Based on this, firstly, by introducing intelligent diagnostic and automatic optimization mechanisms, the reliance on the completeness of user descriptions and system logs can be reduced, improving the accuracy and comprehensiveness of fault identification. Secondly, through user behavior analysis and adaptive threshold adjustment, intelligent identification and dynamic response to complex and novel fault scenarios can be achieved, enhancing the system's adaptability and troubleshooting capabilities. Furthermore, the introduction of bandwidth capacity diagnostic and trend warning functions can identify potential problems in advance, achieving a shift from "passive response" to "proactive prevention." Finally, through a visual interface design, system usability is improved, lowering the operational threshold for non-professional users. This application's embodiments do not rely on image quality or device appearance features, effectively diagnosing non-visually visible internal logic problems, with a wider range of applications and more accurate diagnosis, significantly improving the practicality and intelligence level of self-service troubleshooting.

[0132] This application provides a network optimization device. For example... Figure 9 As shown, the network optimization device 1 includes: Acquisition unit 10 is used to acquire the bandwidth utilization rate of the network interface of the cloud service instance within the current time period; The determining unit 11 is used to determine the threshold corresponding to the current time period based on the seasonal change analysis results and bandwidth usage analysis results of historical bandwidth usage data; and to determine the current bandwidth usage status based on the bandwidth utilization rate and the threshold. Monitoring unit 12 is used to monitor packet loss data on the network interface of the cloud service instance; The determining unit 11 is further configured to determine network optimization information based on the current bandwidth usage status and the packet loss data; the network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment.

[0133] Optionally, the network optimization device further includes: a prediction unit; The determining unit 11 is further configured to determine the first seasonal cycle to which the current time period belongs; determine the first bandwidth baseline value corresponding to the first seasonal cycle based on the seasonal change analysis results; and determine the threshold value based on the first bandwidth baseline value and the bandwidth usage value. The prediction unit is used to predict the bandwidth usage value for the next time period based on the bandwidth usage analysis results.

[0134] Optionally, the network optimization device further includes: an identification unit, a calculation unit, and a fitting unit; The identification unit is used to identify and remove abnormal historical bandwidth usage data from the historical bandwidth usage data to obtain normal historical bandwidth usage data; based on the historical average bandwidth usage data of each time period, it identifies each seasonal period and determines the bandwidth baseline value corresponding to each seasonal period. The calculation unit is used to calculate the historical average bandwidth usage data for each time period based on the normal historical bandwidth usage data. The fitting unit is used to fit the bandwidth usage trend using the historical average bandwidth usage data for each time period as the bandwidth usage analysis result.

[0135] Optionally, the network optimization device further includes: an analysis unit; The acquisition unit 10 is used to acquire the rate limiting policy set on the network interface of the cloud service instance. The analysis unit is used to analyze the speed limiting policy and generate analysis results; the analysis results include, but are not limited to, at least one of the following: the purpose of setting the speed limiting policy, the impact of implementing the speed limiting policy; and generating the cause of packet loss based on the packet loss data and / or the speed limiting policy. The determining unit 11 is further configured to determine the network optimization information based on the packet loss cause, the analysis result, and the current bandwidth usage status.

[0136] Optionally, the determining unit 11 is further configured to: determine whether to switch network paths by performing packet inspection on data packets flowing through the cloud service instance; if a network path switching is determined, identify a problem node based on network topology and network status monitoring information; determine a candidate path from the problem node to the destination node using a shortest path algorithm; the destination node is determined based on the destination address identifier carried by the data packet; determine path evaluation information of the candidate paths, and select a replacement path from the candidate paths based on the path evaluation information; the path evaluation information is determined by at least one of the following: path length, link performance overhead, link load overhead, and historical fault loss indicators; and generate path switching suggestions and / or path switching guidance information based on the replacement path.

[0137] Optionally, the network optimization device further includes: a detection unit; The detection unit is used to detect the first data and generate an alarm message when the first data is detected to be abnormal. The first data includes, but is not limited to, at least one of the following: cloud service instance status, resource status data of resources bound to the cloud service instance, cloud host status data associated with the cloud service instance, order status data of orders related to the cloud service instance, location data of the IP address of the cloud service instance, and blocking status data of the IP address of the cloud service instance.

[0138] Optionally, the network optimization device further includes: a prediction unit; The detection unit is also used to detect the insufficient metering deduction status of orders related to cloud service instances and whether the insufficient metering deduction status has expired. The prediction unit is used to predict the expected metering deduction data for orders related to cloud service instances; The determining unit 11 is further configured to generate a first optimization strategy based on the insufficient metering deduction status, whether the insufficient metering deduction status has expired, and the expected metering deduction data; wherein the first optimization strategy includes, but is not limited to, at least one of the following: adjusting the size of the cloud service instance, or closing unused resources bound to the cloud service instance.

[0139] This application provides a network optimization device that acquires the bandwidth utilization rate of a cloud service instance's network interface within the current time period; determines a threshold corresponding to the current time period based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data; determines the current bandwidth usage status based on the bandwidth utilization rate and the threshold; monitors packet loss data on the cloud service instance's network interface; and determines network optimization information based on the current bandwidth usage status and packet loss data. The network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment. Therefore, the network optimization device proposed in this embodiment dynamically adjusts the bandwidth utilization rate threshold based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data, improving the accuracy of determining the current bandwidth usage status. Combined with packet loss data monitored on the cloud service instance's network interface, it can more comprehensively and accurately determine the causes of network congestion for more complex network problems, thereby providing more accurate network optimization information and improving the flexibility and intelligence of network diagnosis.

[0140] Figure 10 This is a schematic diagram of the composition structure of a network optimization device 1 provided in an embodiment of this application. In practical applications, based on the same disclosed concept of the above embodiments, such as... Figure 10 As shown, the network optimization device 1 in this embodiment includes a processor 13, a memory 14, and a communication bus 15.

[0141] The processor 13 described above can be at least one of the following: Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), CPU, controller, microcontroller, and microprocessor. It is understood that for different devices, the electronic device used to implement the above processor function can also be other types, and this embodiment does not specifically limit it.

[0142] In this embodiment, the communication bus 15 is used to realize the connection communication between the processor 13 and the memory 14; when the processor 13 executes the running program stored in the memory 14, it implements the following network optimization method: Obtain the bandwidth utilization rate of the network interface of the cloud service instance within the current time period; determine the threshold corresponding to the current time period based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data; determine the current bandwidth usage status based on the bandwidth utilization rate and the threshold; monitor packet loss data on the network interface of the cloud service instance; determine network optimization information based on the current bandwidth usage status and the packet loss data; the network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment.

[0143] Furthermore, the processor 13 is also configured to determine the first seasonal cycle to which the current time period belongs; determine the first bandwidth baseline value corresponding to the first seasonal cycle based on the seasonal change analysis results; predict the bandwidth usage value of the next time period based on the bandwidth usage analysis results; and determine the threshold value based on the first bandwidth baseline value and the bandwidth usage value.

[0144] Furthermore, the processor 13 is also configured to identify and remove abnormal historical bandwidth usage data from the historical bandwidth usage data to obtain normal historical bandwidth usage data; calculate the historical average bandwidth usage data for each time period based on the normal historical bandwidth usage data; identify each seasonal period based on the historical average bandwidth usage data for each time period and determine the bandwidth baseline value corresponding to each seasonal period; and fit the bandwidth usage trend using the historical average bandwidth usage data for each time period as the bandwidth usage analysis result.

[0145] Furthermore, the processor 13 is also used to obtain the rate limiting policy set on the network interface of the cloud service instance, analyze the rate limiting policy, and generate analysis results; the analysis results include, but are not limited to, at least one of the following: the purpose of setting the rate limiting policy, the impact of implementing the rate limiting policy; analyze the packet loss data and / or the rate limiting policy to generate the packet loss cause; and determine the network optimization information based on the packet loss cause, the analysis results, and the current bandwidth usage status.

[0146] Furthermore, the processor 13 is also configured to determine whether to switch network paths by performing packet inspection on data packets flowing through the cloud service instance; if a network path switching is determined, identify a problem node based on network topology and network status monitoring information; determine a candidate path from the problem node to the destination node using a shortest path algorithm; the destination node is determined based on the destination address identifier carried by the data packet; determine path evaluation information of the candidate paths, and select a replacement path from the candidate paths based on the path evaluation information; the path evaluation information is determined by at least one of the following: path length, link performance overhead, link load overhead, and historical fault loss indicators; and generate path switching suggestions and / or path switching guidance information based on the replacement path.

[0147] Furthermore, the processor 13 is also used to detect the first data and generate an alarm message when the first data is detected to be abnormal. The first data includes, but is not limited to, at least one of the following: cloud service instance status, resource status data of resources bound to the cloud service instance, cloud host status data associated with the cloud service instance, order status data of orders related to the cloud service instance, location data of the IP address of the cloud service instance, and blocking status data of the IP address of the cloud service instance.

[0148] Furthermore, the processor 13 is also used to detect the insufficient metering deduction status of cloud service instance related orders and whether the insufficient metering deduction status has expired; predict the expected metering deduction data of cloud service instance related orders; and generate a first optimization strategy based on the insufficient metering deduction status, whether the insufficient metering deduction status has expired, and the expected metering deduction data; wherein the first optimization strategy includes, but is not limited to, at least one of the following: adjusting the size of the cloud service instance and closing unused resources bound to the cloud service instance.

[0149] This application provides a storage medium storing a computer program thereon. The computer-readable storage medium stores one or more programs, which can be executed by one or more processors and applied in a network optimization device. The computer program implements the network optimization method described above.

[0150] Based on the above embodiments, this application provides a computer program product, including a computer program that can be executed by one or more processors, and the computer program implements the network optimization method described above.

[0151] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0152] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause an image display device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.

[0153] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A network optimization method, characterized in that, The method includes: Get the bandwidth utilization rate of the network interface of the cloud service instance in the current time period; The threshold corresponding to the current time period is determined based on the seasonal variation analysis results of historical bandwidth usage data and the bandwidth usage analysis results; and the current bandwidth usage status is determined based on the bandwidth utilization rate and the threshold. Monitor packet loss data on the network interfaces of cloud service instances; Based on the current bandwidth usage status and the packet loss data, network optimization information is determined; the network optimization information includes, but is not limited to, at least one of the following: traffic optimization, network configuration optimization, rate limiting policy adjustment, and bandwidth capacity adjustment.

2. The method according to claim 1, characterized in that, The step of determining the threshold corresponding to the current time period based on the seasonal variation analysis results and bandwidth usage analysis results of historical bandwidth usage data includes: Determine the first seasonal cycle to which the current time period belongs; Based on the seasonal variation analysis results, the first bandwidth baseline value corresponding to the first seasonal cycle is determined; Based on the bandwidth usage analysis results, predict the bandwidth usage value for the next time period; The threshold is determined based on the first bandwidth baseline value and the bandwidth usage value.

3. The method according to claim 1 or 2, characterized in that, The method further includes: Abnormal historical bandwidth usage data is identified and removed from the historical bandwidth usage data to obtain normal historical bandwidth usage data. Calculate the historical average bandwidth usage data for each time period based on the normal historical bandwidth usage data; Based on the historical average bandwidth usage data for each time period, identify each seasonal period and determine the corresponding bandwidth baseline value for each seasonal period. By using historical average bandwidth usage data for each time period, a bandwidth usage trend is fitted as the result of the bandwidth usage analysis.

4. The method according to claim 1, characterized in that, After monitoring packet loss data on the network interface of the cloud service instance; before determining network optimization information based on the current bandwidth usage status and the packet loss data, the method further includes: Obtain the rate limiting policy set on the network interface of the cloud service instance, analyze the rate limiting policy, and generate analysis results; the analysis results include, but are not limited to, at least one of the following: the purpose of setting the rate limiting policy and the impact of implementing the rate limiting policy; Based on the packet loss data and / or the rate limiting strategy, the cause of packet loss is generated. The step of determining network optimization information based on the current bandwidth usage status and the packet loss data includes: Based on the reasons for packet loss, the analysis results, and the current bandwidth usage status, the network optimization information is determined.

5. The method according to claim 1, characterized in that, The method further includes: By performing packet inspection on the data packets flowing through the cloud service instance, it can be determined whether to switch network paths; Once the network path to be switched is determined, the problem node is identified based on the network topology and network status monitoring information; A shortest path algorithm is used to determine a candidate path from the problem node to the destination node; the destination node is determined based on the destination address identifier carried in the data packet. Determine the path evaluation information of the candidate paths, and select a replacement path from the candidate paths based on the path evaluation information; the path evaluation information is determined by at least one of the following: path length, link performance overhead, link load overhead, and historical failure loss index. Based on the replacement path, generate path switching suggestions and / or path switching guidance information.

6. The method according to claim 1, characterized in that, The method further includes: The first data is inspected, and an alarm message is generated when an anomaly is detected in the first data. The first data includes, but is not limited to, at least one of the following: cloud service instance status, resource status data of resources bound to the cloud service instance, cloud host status data associated with the cloud service instance, order status data of orders related to the cloud service instance, location data of the Internet Protocol IP address of the cloud service instance, and blocking status data of the IP address of the cloud service instance.

7. The method according to claim 1, characterized in that, The method further includes: Detect the insufficient metering deduction status of orders related to cloud service instances and whether the insufficient metering deduction status has expired; Predicted metering deduction data for orders related to cloud service instances; Based on the insufficient metering deduction status, whether the insufficient metering deduction status has expired, and the expected metering deduction data, a first optimization strategy is generated; wherein, the first optimization strategy includes, but is not limited to, at least one of the following: adjusting the size of the cloud service instance, or disabling unused resources bound to the cloud service instance.

8. A network optimization device, characterized in that, The network optimization device includes a processor and a memory; when the processor executes the running program stored in the memory, it implements the method as described in any one of claims 1-7.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, The computer program, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images