Method, device and equipment for adjusting network load, medium and product

By obtaining the operating status data of each resource pool of the network system and applying a proportional integral differential control algorithm, the load sharing ratio is dynamically adjusted, which solves the problems of insufficient load balancing and link congestion in computing power network routing, and improves system performance and resource utilization rate .

CN119030983AInactive Publication Date: 2024-11-26CHINA MOBILE COMM LTD RES INST +1
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411063641.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing computing power network routing is limited in large-scale and dynamic environments, insufficient load balancing, and high risk of link congestion, especially in asymmetric networks, which affects system performance.

Method used

By obtaining the operating status data of each resource pool of the network system, adjusting the load sharing ratio based on the proportional integral differential control algorithm, dynamically adjusting traffic allocation to ensure load balancing and effective resource utilization.

Benefits of technology

It improves the overall performance and reliability of the network system, avoids load imbalance and link congestion, and improves resource utilization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119030983A_ABST
    Figure CN119030983A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of computing power networks, and provides a method, a device, equipment, a medium and a product for adjusting a network load, and the method comprises the steps: obtaining the operation state data of each resource pool in a network system; determining a load distribution strategy according to the running state data; and the load sharing proportion of each resource pool is adjusted based on the load distribution strategy, and the load sharing proportion is the flow proportion of the data flow of the service in the network system shared by each resource pool. Therefore, the load sharing proportion of each resource pool can be adjusted based on the operation state data of each resource pool, the problem of ineffective utilization of network resources or link congestion caused by load imbalance is avoided, the overall performance and reliability of a network system are improved, and the utilization rate of network links and network resources is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computing power network technology, and in particular to a method, device, equipment, medium and product for adjusting network load. Background Art

[0002] At present, computing power has become a vital digital industry infrastructure. The sinking of computing power to the edge and even the super edge has become an industry trend. As one of the key technologies of computing power network, computing power routing can be generally divided into centralized computing power routing solutions and distributed computing power routing solutions.

[0003] Among them, the centralized computing power routing solution based on IPv6+ (Internet Protocol Version 6+, an upgrade of Internet Protocol Version 6) technology can flexibly dispatch computing power, but its scalability is limited in large-scale and dynamic environments. In distributed computing power routing, the existing IGP (Interior Gateway Protocol) and BGP (Border Gateway Protocol) are extended to support distributed computing power routing. Although it can increase the transmission efficiency of computing power information, it also increases the complexity of network management and configuration. In addition, the current computing power resources in the cloud are usually in a closed mode, which limits the flexible sharing and dynamic configuration of resources, and is not conducive to the rapid deployment and adjustment of services.

[0004] LB (Load Balance) is an essential key component of high-concurrency and high-availability systems. Its goal is to try to evenly distribute network traffic to multiple servers to improve the overall response speed and availability of the system. There are usually three types of load balancing: flow-based load balancing, packet-based load balancing, and bandwidth-based unbalanced load balancing. Among them, flow-based or packet-based load balancing solutions such as ECMP (Equal-Cost Multi-Path Routing) have the risk of link congestion because they have no congestion awareness mechanism, which may cause the already congested paths to further increase the burden. Moreover, in an asymmetric network, some paths have higher bandwidth or lower latency, and ECMP fails to effectively utilize these advantageous paths, resulting in uneven load distribution. And when the size of the flow varies greatly, ECMP may cause some traffic to be concentrated on a few paths, resulting in a significant decrease in the effect of flow-based load balancing, affecting the overall performance of the system. Summary of the invention

[0005] The embodiments of the present application provide a method, device, equipment, medium and product for adjusting network load to solve the technical problems of insufficient load balancing and link congestion risks faced by computing power network routing in related technologies.

[0006] To solve the above technical problems, the embodiments of the present application provide the following aspects:

[0007] In a first aspect, an embodiment of the present application provides a method for adjusting a network load, including:

[0008] Obtain the operating status data of each resource pool in the network system;

[0009] Determine a load distribution strategy according to the operating status data;

[0010] The load sharing ratio of each resource pool is adjusted based on the load distribution strategy, wherein the load sharing ratio is the traffic ratio of the service data flow in the network system shared by each resource pool.

[0011] Optionally, the operating status data includes at least one of the following: response time of the resource pool, data throughput of the resource pool, memory usage of the resource pool, and number of connections of resources in the resource pool.

[0012] Optionally, obtaining the operation status data of each resource pool in the network system includes:

[0013] The operation status data of each resource pool in the network system is obtained at a preset period, wherein the preset period is consistent with the update frequency of the border gateway protocol routing information of the network system.

[0014] Optionally, obtaining the operation status data of each resource pool in the network system includes:

[0015] The operation status data of each resource pool in the network system is obtained through the interface of the border gateway protocol and / or the presentation layer state transfer architecture style.

[0016] Optionally, determining a load distribution strategy according to the operating status data includes:

[0017] Outputting a control signal based on a proportional-integral-derivative control algorithm according to the operating status data and the average operating status data of each resource pool;

[0018] determining the load distribution strategy based on the control signal;

[0019] Among them, the input of the proportional-integral-differential control algorithm is the average operating status data and the error value of the operating status data; the control signal is composed of the proportional part of the error value, the integral part of the error value and the differential part of the error value, and the control signal is used to adjust the load sharing ratio so that the operating status data is close to the average operating status data.

[0020] Optionally, outputting a control signal based on a proportional-integral-derivative control algorithm according to the operating status data and the average operating status data of each resource pool includes:

[0021] determining the error value between the operating status data and the average operating status data;

[0022] Based on the error value and a preset proportional gain coefficient, a preset integral gain coefficient and a preset differential gain coefficient in the proportional-integral-differential control algorithm, the proportional part, the integral part and the differential part are determined respectively;

[0023] The sum of the proportional part, the integral part and the differential part is determined as the control signal, and the control signal is output.

[0024] Optionally, adjusting the load sharing ratio of each resource pool based on the load distribution strategy includes:

[0025] Regulating the traffic ratio of the service data flow in the locally stored routing table based on the load distribution strategy to obtain the regulated traffic ratio;

[0026] Based on the regulated traffic ratio, the load sharing ratio of each resource pool is determined.

[0027] Optionally, regulating the flow ratio of the service data flow in the locally stored routing table based on the load distribution strategy, and obtaining the regulated flow ratio includes:

[0028] Determine a target resource pool from among the resource pools, and perform at least one of the following steps:

[0029] If the load distribution strategy is used to indicate increasing the load sharing ratio of the target resource pool, then increasing the traffic ratio corresponding to the target resource pool in the routing table to obtain the regulated traffic ratio;

[0030] If the load distribution strategy is used to indicate reducing the load sharing ratio of the target resource pool, then reducing the traffic ratio corresponding to the target resource pool in the routing table to obtain the regulated traffic ratio;

[0031] If the load distribution strategy is used to indicate that the load sharing ratio of the target resource pool remains unchanged, then the traffic ratio corresponding to the target resource pool in the routing table is maintained unchanged to obtain the regulated traffic ratio.

[0032] In a second aspect, an embodiment of the present application provides a device for adjusting a network load, including:

[0033] An acquisition module is used to obtain the operation status data of each resource pool in the network system;

[0034] An execution module, configured to determine a load distribution strategy according to the operation status data;

[0035] The load sharing ratio of each resource pool is adjusted based on the load distribution strategy, wherein the load sharing ratio is the traffic ratio of the service data flow in the network system shared by each resource pool.

[0036] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of a method for adjusting network load as described in the first aspect are implemented.

[0037] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for adjusting network load as described in the first aspect are implemented.

[0038] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising computer instructions, which, when executed by a processor, implement the steps of a method for adjusting network load as described in the first aspect.

[0039] Therefore, the load sharing ratio of each resource pool can be adjusted based on the operating status data of each resource pool, avoiding the problem of ineffective utilization of network resources or link congestion caused by load imbalance, improving the overall performance and reliability of the network system, and improving the utilization rate of network links and network resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0041] Figure 1 A flowchart of a method for adjusting network load provided in an embodiment of the present application;

[0042] Figure 2 A block diagram of a computing power network architecture provided for an embodiment of the present application;

[0043] Figure 3 A flowchart of a method for adjusting network load provided in an embodiment of the present application;

[0044] Figure 4A flowchart of a method for adjusting network load provided in an embodiment of the present application;

[0045] Figure 5 A structural block diagram of a device for adjusting network load provided in an embodiment of the present application;

[0046] Figure 6 A structural block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0048] Figure 1 A method for adjusting network load provided in accordance with an embodiment of the present application is shown, such as Figure 1 As shown, including:

[0049] Step S101: Acquire the operation status data of each resource pool in the network system;

[0050] Step S102: Determine a load distribution strategy according to the operation status data;

[0051] Step S103: adjusting the load sharing ratio of each resource pool based on the load distribution strategy;

[0052] The load sharing ratio refers to the traffic ratio of the service data flow in the network system shared by each resource pool.

[0053] It should be noted that Figure 1 The method shown can be applied to an embedded device (entry device) that is integrated into a router or a switch to implement intelligent and automated load sharing control. Figure 1 The method shown can be applied to Figure 2 In the network architecture shown in Figure 2 As shown, the method shown in the embodiment of the present application is mainly used for load balancing on the ingress router. When the customer traffic reaches the ingress router, load balancing can be performed on the ingress router according to the load balancing algorithm, so that the customer traffic can be dispatched to the cloud resource pool in the best situation. A component or algorithm module can be added to the ingress router to dynamically adjust the load sharing ratio of each resource pool to achieve the best performance.

[0054] In step S101, the operation status data of each resource pool in the network system can be obtained through the BGP protocol and / or the Restful interface (i.e., the interface of the presentation layer state transfer architecture style). Figure 2 In the network architecture shown, the cloud management platform can collect the operating status data of each resource pool in the network system and pass it to the computing power router (entrance), that is, the embedded device, through the Restful interface. The computing power router (exit) can collect the operating status data of each resource pool in the network system and pass it to the computing power router (entrance), that is, the embedded device, through the BGP protocol.

[0055] It should be noted that when collecting data, sensors and monitoring tools can be deployed on key nodes (such as servers and routers) to collect real-time load data. SNMP (Simple Network Management Protocol) can be used to regularly pull device performance data. Time series databases or big data platforms can be used to store and process large amounts of load data. Technologies such as data cleaning and normalization can be applied to ensure data quality. Stream processing technologies (such as Apache Kafka) can also be used to analyze real-time data to quickly respond to changes in system load.

[0056] In step S101, a resource pool refers to a set of sharable computing, storage and network resources that can be dynamically allocated and used by multiple users or applications. The concept of a resource pool is usually closely related to virtualization and cloud computing. A resource pool usually includes: computing resources, including the server's CPU (Central Processing Unit), memory and processing power. Computing resources can be allocated through virtualization technology so that multiple virtual machines can run on the same physical server; storage resources, including hard disks, solid-state drives and network storage devices, can be allocated on demand to support the data storage needs of different applications. Network resources, including bandwidth, IP (Internet Protocol) addresses, network devices (such as switches and routers), etc. Network resources can be managed and allocated through virtual network technology to achieve efficient data transmission.

[0057] The operation status data includes at least one of the following: response time of the resource pool, data throughput of the resource pool, memory usage of the resource pool, and number of connections of resources in the resource pool.

[0058] In a possible implementation, step S101, obtaining the operating status data of each resource pool in the network system includes: obtaining the operating status data of each resource pool in the network system at every preset period, wherein the preset period is consistent with the update frequency of the Border Gateway Protocol BGP routing information of the network system.

[0059] It should be noted that BGP is a routing protocol used to propagate routing information between different autonomous systems, and BGP routing information update is a mechanism used by BGP to notify neighboring systems of routing changes. BGP routing information update is part of network operation, allowing network devices (such as routers) to know how to effectively direct data to their destination. When the network path changes (such as link failure, policy adjustment, new link activation), the relevant router will send an update to notify other routers of these changes. The frequency of BGP routing information updates depends on the dynamic nature of the network environment. In a stable network environment, updates may be less; while in a dynamically changing environment, such as frequent changes in routing policies or adjustments to network structures, updates may be more frequent. In the method shown in the embodiment of the present application, the preset period is kept consistent with the update frequency of the BGP routing information of the network system, which has the following effects: it can respond to the latest changes in the network status in real time, which is very important for making decisions and adjustments based on the latest network conditions, especially in application scenarios that require dynamic allocation of network resources or load balancing. If the sampling frequency is much lower than the frequency of BGP routing information updates, the controller may make decisions based on outdated data, thereby reducing control effects and system performance. If the sampling frequency is too high, it may cause unnecessary burden on the system and increase processing complexity without corresponding performance improvement. Keeping consistent with the BGP routing information update frequency can prevent system instability caused by excessive response. Excessively frequent responses may cause oscillation of control signals, especially when network status changes rapidly. A reasonable sampling period can balance the system's response time and the use of computing resources, achieving efficient network management and optimizing resource allocation.

[0060] In one possible implementation, Figure 3 As shown, step S102, determining a load distribution strategy according to the operating status data, includes:

[0061] Step S301: Output a control signal based on a PID control algorithm according to the operation status data and the average operation status data of each resource pool;

[0062] Step S302: Determine a load distribution strategy based on the control signal.

[0063] In a possible implementation, outputting a control signal based on a proportional-integral-differential control algorithm according to the operating status data and the average operating status data of each resource pool includes:

[0064] determining an error value between the operating status data and the average operating status data;

[0065] Based on the error value and the preset proportional gain coefficient, the preset integral gain coefficient, and the preset differential gain coefficient in the proportional-integral-differential control algorithm, the proportional part, the integral part, and the differential part are determined respectively;

[0066] The sum of the proportional part, the integral part and the differential part is determined as a control signal, and the control signal is output.

[0067] Among them, the input of the PID (Proportional-Integral-Derivative) control algorithm is the average operating status data and the error value of the operating status data; the control signal is composed of the proportional part of the error value, the integral part of the error value and the differential part of the error value. The control signal is used to adjust the load sharing ratio so that the operating status data is close to the average operating status data.

[0068] It should be noted that the PID controller is a widely used control algorithm for adjusting the output of a system to match a desired reference input. The PID controller is simple, reliable, and has shown good performance in many different applications. The basic principle of the PID controller is that the PID controller calculates a control signal based on the error e(t) = r(t) - y(t) between the reference input r(t) and the actual output y(t) of the system. This control signal is then used to adjust the behavior of the system (in the method shown in the embodiment of the present application, it refers to the network system).

[0069] The output of the PID controller, i.e. the control signal u(t), consists of the following three parts:

[0070] 1. Proportional (P) part: directly proportional to the error. If the error is large, the control signal is also large, which helps the system respond to the error quickly. However, pure proportional control may cause system instability, especially when the error is large.

[0071] P = Kp*e(t), where Kp is the proportional coefficient, e(t) is the error, and P represents the proportional part.

[0072] 2. Integral (I) part: The integral of the error. The integral part helps to eliminate steady-state errors, but may increase the response time of the system.

[0073] Among them, Ki is the integration coefficient, e(t) is the error, and I represents the integral.

[0074] 3. Differential (D) part: The derivative of the error. The differential part can predict the future changes of the error and adjust the control signal in advance accordingly, thereby improving the dynamic performance of the system.

[0075] D=Kd*de(t) / dt, where Kd is the differential coefficient, e(t) is the error, and D represents the differential.

[0076] The total output of the PID controller is a weighted sum of the three components above:

[0077]

[0078] Among them, Kp is the proportional gain, Ki is the integral gain, Kd is the differential gain, and e(t) is the error, which refers to the error between the average load of each resource pool and the load of the currently controlled resource pool (that is, the difference between the operating status data in step S201 and the average operating status data of each resource pool). The performance of the PID controller depends on the selection of these three parameters (Kp, Ki and Kd).

[0079] In the method shown in the embodiment of the present application, the PID controller is mainly used for the problem of computing network load balancing. The control goal is to make the relative load of all computing servers the same. Therefore, the control deviation is the difference between the relative load of each computing server and the average relative load, as shown in the following objective function: Where V i (t) represents the relative load of computing resource pool i at the current time t, It indicates the average relative load of all computing power at the current moment. Through experimental comparison, it is found that when the sampling period is 30S (consistent with the BGP route update frequency, the determining factor for selecting the sampling period is the update frequency of computing power information, or the frequency of route changes), Kp is 1.1, Ki is 0.04, and Kd is 0.08, the algorithm converges fastest and the load sharing is most balanced.

[0080] However, 30S, 1.1, 0.04 and 0.08 are only used as an exemplary description. The specific numerical settings can be selected and adjusted according to actual needs, that is, the network system can regularly collect the current load and resource status information, provide necessary feedback for the next control cycle, and adjust the PID parameters according to the feedback data to optimize the control strategy to cope with changes in network status. Specifically, key performance indicators (such as CPU usage, memory usage, network traffic, etc.) can be collected in real time through monitoring tools, and these data can be analyzed to identify system load patterns and change trends, and the initial PID parameters can be determined using experiments or simulation tests, and then the PID parameters can be dynamically adjusted according to real-time data feedback to optimize system response. Adjusting these parameters is a key step, and it is usually necessary to iterate and adjust based on the characteristics of the system and the desired behavior. Generally, these parameters can be adjusted using rules of thumb, trial and error, or more advanced optimization techniques (such as genetic algorithms, particle swarm optimization, etc.). For example, if the system load changes rapidly, the differential gain Kd may need to be increased to reduce overshoot and oscillation.

[0081] Three different load sharing algorithms were used for testing:

[0082] Global optimal algorithm: directly select the computing power resource with the smallest negative metric (computing power) value;

[0083] ECMP algorithm: round-robin selection is performed in a 1:1:1:1 manner;

[0084] PID algorithm: Based on the resource and load status information periodically fed back by the server, the percentage of newly arrived services going to different servers is refreshed. When the first packet arrives, the load is shared according to the percentage in the table entry.

[0085] The test comparison results are shown in Table 1, which is a load sharing performance test table of the global optimal algorithm, ECMP algorithm and PID algorithm:

[0086] Table 1

[0087]

[0088] The PID algorithm can dynamically improve resource utilization and increase business call volume. Compared with other algorithms, the PID algorithm has obvious advantages in convergence speed and convergence effect. Compared with the ECMP algorithm, the PID algorithm achieves about 20% improvement in computing resource availability through load balancing. Compared with the global optimal algorithm, the PID algorithm achieves about 80% of the network-side computing attribute transmission burden. Therefore, the introduction of the PID control algorithm into the router can dynamically adjust the traffic sharing ratio in different resource pools to achieve optimal performance, improve resource utilization and system stability.

[0089] In one possible implementation, step 103, adjusting the load sharing ratio of each resource pool based on the load distribution strategy, includes: adjusting the traffic ratio of the business data flow in the locally stored routing table based on the load distribution strategy to obtain the adjusted traffic ratio; based on the adjusted traffic ratio, determining the load sharing ratio of each resource pool.

[0090] That is to say, the specific regulation process can be based on the routing table. First, the table items in the routing table are changed, and then the load sharing ratio of each resource pool is adjusted based on the changed table items in the routing table.

[0091] In a specific application scenario, after the inlet routing node receives the computing power routing transmitted from one or more egress routing nodes, it can generate a local computing power routing table according to the local preset load sharing algorithm, as shown in Table 2, which is the initial computing power routing table. And the value of the percentage table item in the initial computing power routing table can be adjusted based on the load distribution strategy, so as to adjust the load sharing ratio of each resource pool based on the value of the percentage table item.

[0092] Table 2

[0093]

[0094] It should be noted that the PID control module can be integrated inside the router and switch, directly connected to the routing decision mechanism of the device, and the computing power routing table can be dynamically adjusted according to the PID control signal to optimize traffic distribution. The computing power routing table contains the available computing power and priority information of each resource pool to ensure that traffic is routed according to performance optimization.

[0095] In one possible implementation, the traffic ratio of the business data flow in the locally stored routing table is regulated based on the load distribution strategy to obtain the regulated traffic ratio, including: determining a target resource pool in each resource pool, and executing at least one of the following steps: if the load distribution strategy is used to indicate an increase in the load sharing ratio of the target resource pool, then the traffic ratio corresponding to the target resource pool in the routing table is increased to obtain the regulated traffic ratio; if the load distribution strategy is used to indicate a decrease in the load sharing ratio of the target resource pool, then the traffic ratio corresponding to the target resource pool in the routing table is reduced to obtain the regulated traffic ratio; if the load distribution strategy is used to indicate that the load sharing ratio of the target resource pool remains unchanged, then the traffic ratio corresponding to the target resource pool in the routing table is maintained unchanged to obtain the regulated traffic ratio.

[0096] In a specific application scenario, taking the computing power routing table shown in Table 2 as an example, the specific load sharing ratio is determined by collecting the real-time traffic and load data of each resource pool (R2, R3, R4), and calculating the error value between the current load and the target load. For example, if the current load of R2 is 50%, and the target load is 40%, and the error is 10%, then it is necessary to reduce the traffic of the new business to R2, and accordingly, the traffic of the new business to R3 and R4 can be increased. Based on the PID control algorithm, a specific control signal can be calculated, and the load distribution strategy is determined based on the control signal to reduce the traffic of the new business to R2 (i.e., reduce the load sharing ratio of the target resource pool). Then, the corresponding percentage of R2 can be adjusted from 40% to 30%, and based on the adjusted traffic ratio (i.e., 30%), the load sharing ratio of R2 can be determined to be 30%.

[0097] Therefore, network load balancing based on PID control can be effectively achieved, improving system performance and reliability. It can ensure that the network maintains optimal performance under dynamically changing load conditions while ensuring data security and real-time performance.

[0098] In summary, the method shown in the embodiment of the present application can optimize the design of a load sharing algorithm based on the PID algorithm, and can add a component or module to the router device, and periodically collect the business indicators of the resource pool to adjust the target index. The target index includes but is not limited to response time, throughput, CPU utilization, memory utilization, number of resource connections, GPU (Graphics Processing Unit), resource utilization, site preference settings, etc., which are used for computing power routing traffic tuning to avoid system load overload, while optimizing computing power network load and improving business experience.

[0099] The method shown in the embodiment of the present application adopts multi-objective optimization, taking into account multiple objective functions, such as response time, throughput, CPU utilization, memory utilization, number of resource connections, GPU resource utilization, site preference settings, etc., and uses optimization algorithms to achieve multi-objective optimization to obtain better performance.

[0100] The method shown in the embodiment of the present application can embed a PID control unit (control module) on a router, a switch, and a load sharing device to control the completion of computing power routing traffic tuning, and generate a computing power routing table on the device, and then perform load tuning based on the computing power routing table to improve the overall performance and reliability of the network system and improve the utilization rate of network links and network resources.

[0101] An embedded device can be designed according to the method shown in the embodiment of the present application, and integrated into a router or switch to realize intelligent automatic control load sharing. The entire workflow is as follows: Figure 4 As shown, the whole process is divided into 5 steps:

[0102] Collect system operation data, including current load, system status, business call volume, or hardware resource utilization rate;

[0103] Use BGP protocol or RESTful interface to periodically notify the ingress device of the information;

[0104] The ingress device selects a load balancing algorithm and performs load balancing based on flows, distributing the services among different resource pools.

[0105] The business call beam or load condition is fed back to the entry as a PID control quantity to adjust the load algorithm.

[0106] Figure 4 The steps shown are a continuous cycle. By continuously collecting data, adjusting parameters, and evaluating results, the load of all resource pools can be relatively balanced and the business call volume can be increased.

[0107] Figure 5 A device for adjusting network load according to an embodiment of the present application is shown, such as Figure 5 As shown, the device 50 includes:

[0108] The acquisition module 501 is used to acquire the operation status data of each resource pool in the network system;

[0109] Execution module 502 is used to determine the load distribution strategy according to the operation status data; adjust the load sharing ratio of each resource pool based on the load distribution strategy, wherein the load sharing ratio is the traffic ratio of the business data flow shared by each resource pool in the network system.

[0110] In a possible implementation, the operation status data includes at least one of the following: a response time of the resource pool, a data throughput of the resource pool, a memory usage rate of the resource pool, and a number of connections of resources in the resource pool.

[0111] In a possible implementation, the acquisition module 501 is further configured to acquire the operation status data of each resource pool in the network system at every preset period, wherein the preset period is consistent with the update frequency of the border gateway protocol routing information of the network system.

[0112] In a possible implementation, the acquisition module 501 is further configured to acquire the operation status data of each resource pool in the network system through a border gateway protocol and / or a presentation state transfer architecture style interface.

[0113] In a possible implementation, the execution module 502 is further configured to output a control signal based on a proportional-integral-derivative control algorithm according to the operation status data and the average operation status data of each resource pool; and determine a load distribution strategy based on the control signal;

[0114] Among them, the input of the proportional-integral-differential control algorithm is the average operating status data and the error value of the operating status data; the control signal is composed of the proportional part of the error value, the integral part of the error value and the differential part of the error value, and the control signal is used to adjust the load sharing ratio so that the operating status data is close to the average operating status data.

[0115] In a possible implementation, the execution module 502 is further used to determine an error value between the operating state data and the average operating state data; based on the error value and a preset proportional gain coefficient, a preset integral gain coefficient, and a preset differential gain coefficient in the proportional-integral-differential control algorithm, the proportional part, the integral part, and the differential part are determined respectively;

[0116] The sum of the proportional part, the integral part and the differential part is determined as a control signal, and the control signal is output.

[0117] In a possible implementation, the execution module 502 is further configured to adjust the flow ratio of the service data flow in the locally stored routing table based on the load distribution strategy to obtain the adjusted flow ratio;

[0118] Based on the adjusted traffic ratio, determine the load sharing ratio of each resource pool.

[0119] In one possible implementation, the execution module 502 is also used to determine the target resource pool in each resource pool, and execute at least one of the following steps: if the load distribution strategy is used to indicate an increase in the load sharing ratio of the target resource pool, then the traffic ratio corresponding to the target resource pool in the routing table is increased to obtain the regulated traffic ratio; if the load distribution strategy is used to indicate a decrease in the load sharing ratio of the target resource pool, then the traffic ratio corresponding to the target resource pool in the routing table is reduced to obtain the regulated traffic ratio; if the load distribution strategy is used to indicate that the load sharing ratio of the target resource pool remains unchanged, then the traffic ratio corresponding to the target resource pool in the routing table is maintained unchanged to obtain the regulated traffic ratio.

[0120] The device for adjusting network load shown in the embodiment of the present application proposes a load sharing algorithm for automatically controlling PID, which can be integrated in a router or switch for load sharing to complete the scheduling and optimization of computing power routing traffic. And it has the following technical effects:

[0121] It can overcome the shortcomings of traditional load sharing, such as the possible increase in link congestion, poor use of asymmetric networks, and poor flow-based load balancing. It can dynamically and in real time select the path that best meets business needs based on the user's network needs and computing power requirements in the network, improve resource utilization, and achieve a green and energy-saving network.

[0122] It has certain application value, can provide efficient deployment of services, has broad application prospects, can be applied in network environments such as LAN and WAN, can select the path that best meets business needs according to the user's network needs and computing power needs, improve resource utilization, and realize green and energy-saving networks. It can also provide intelligent and efficient load sharing algorithms for network load sharing services to improve link and resource utilization.

[0123] The present application embodiment also provides an electronic device 60, such as Figure 6 As shown, it includes: a processor 601, a memory 602, and a program stored in the memory 602 and executable on the processor 601. When the program is executed by the processor, the steps of a method for adjusting network load as shown in the above embodiment are implemented.

[0124] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of a method for adjusting the network load shown in the above method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0125] An embodiment of the present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of a method for adjusting network load shown in the above method embodiment and can achieve the same technical effect. To avoid repetition, they are not repeated here.

[0126] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0127] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0128] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A method for adjusting network load, characterized in that: include: Obtain the operating status data of each resource pool in the network system; Determine a load distribution strategy according to the operating status data; The load sharing ratio of each resource pool is adjusted based on the load distribution strategy, wherein the load sharing ratio is the traffic ratio of the service data flow in the network system shared by each resource pool.

2. The method according to claim 1, characterized in that The operation status data includes at least one of the following: the response time of the resource pool, the data throughput of the resource pool, the memory usage of the resource pool, and the number of connections of the resources in the resource pool.

3. The method according to claim 1, characterized in that Obtaining the operating status data of each resource pool in the network system includes: The operation status data of each resource pool in the network system is obtained at a preset period, wherein the preset period is consistent with the update frequency of the border gateway protocol routing information of the network system.

4. The method according to claim 1, characterized in that Obtaining the operating status data of each resource pool in the network system includes: The operation status data of each resource pool in the network system is obtained through the interface of the border gateway protocol and / or the presentation layer state transfer architecture style.

5. The method according to claim 1, characterized in that: Determining a load distribution strategy according to the operating status data includes: Outputting a control signal based on a proportional-integral-derivative control algorithm according to the operating status data and the average operating status data of each resource pool; determining the load distribution strategy based on the control signal; Among them, the input of the proportional-integral-differential control algorithm is the average operating status data and the error value of the operating status data; the control signal is composed of the proportional part of the error value, the integral part of the error value and the differential part of the error value, and the control signal is used to adjust the load sharing ratio so that the operating status data is close to the average operating status data.

6. The method according to claim 5, characterized in that Outputting a control signal based on a proportional-integral-differential control algorithm according to the operating status data and the average operating status data of each resource pool includes: determining the error value between the operating status data and the average operating status data; Based on the error value and a preset proportional gain coefficient, a preset integral gain coefficient and a preset differential gain coefficient in the proportional-integral-differential control algorithm, the proportional part, the integral part and the differential part are determined respectively; The sum of the proportional part, the integral part and the differential part is determined as the control signal, and the control signal is output.

7. The method according to claim 1, characterized in that Adjusting the load sharing ratio of each resource pool based on the load distribution strategy includes: Regulating the traffic ratio of the service data flow in the locally stored routing table based on the load distribution strategy to obtain the regulated traffic ratio; Based on the regulated traffic ratio, the load sharing ratio of each resource pool is determined.

8. The method according to claim 7, characterized in that The flow ratio of the service data flow in the locally stored routing table is regulated based on the load distribution strategy, and the regulated flow ratio is obtained including: Determine a target resource pool from among the resource pools, and perform at least one of the following steps: If the load distribution strategy is used to indicate increasing the load sharing ratio of the target resource pool, then increasing the traffic ratio corresponding to the target resource pool in the routing table to obtain the regulated traffic ratio; If the load distribution strategy is used to indicate reducing the load sharing ratio of the target resource pool, then reducing the traffic ratio corresponding to the target resource pool in the routing table to obtain the regulated traffic ratio; If the load distribution strategy is used to indicate that the load sharing ratio of the target resource pool remains unchanged, then the traffic ratio corresponding to the target resource pool in the routing table is maintained unchanged to obtain the regulated traffic ratio.

9. A device for adjusting network load, characterized in that: include: An acquisition module is used to obtain the operation status data of each resource pool in the network system; An execution module, configured to determine a load distribution strategy according to the operation status data; The load sharing ratio of each resource pool is adjusted based on the load distribution strategy, wherein the load sharing ratio is the flow ratio of the data flow of the service in the network system shared by each resource pool.

10. An electronic device, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of a method for adjusting network load as described in any one of claims 1 to 6 are implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of a method for adjusting a network load according to any one of claims 1 to 6 are implemented.

12. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of a method for adjusting network load according to any one of claims 1 to 6.

Citation Information

Patent Citations

  • Load balancing method, apparatus, computer device, and storage medium

    CN109067926A

  • Resource allocation method and system based on big data

    CN116467082A

  • Distributed energy resource cooperative regulation and control method for power load regulation and control terminal

    CN118117583A

  • Method, device, system and equipment for scheduling computing power network resources and storage medium

    CN118394513A

  • Computing power load balancing method and apparatuses

    WO2024007171A1