Distributed database load dynamic routing method, device and equipment and storage medium

Abstract

The invention relates to a distributed database load dynamic routing method, device and equipment and a storage medium, and the method comprises the steps: obtaining the information of a to-be-processed load and the state information of all computing nodes in a cluster, and obtaining the current load level information of each computing node according to the state information of the computing nodes; inputting the information of the to-be-processed load into the load resource model to obtain the predicted resource occupation amount of each computing node for processing the to-be-processed load, and according to the current load level information of each computing node and the predicted resource occupation amount for processing the to-be-processed load, calculating the to-be-processed load according to the predicted resource occupation amount. And selecting the computing node with the lowest predicted load level to process the to-be-processed load. According to the technical scheme, the to-be-processed load can be dynamically routed to the computing node with the lowest resource occupation amount when the load is processed for processing, the load of each computing node in the cluster is effectively balanced, the probability that the computing nodes are abnormal due to load imbalance is reduced, and the computing resources of the computing nodes are fully utilized.

Description

Technical Field

[0001] The present application relates to the field of distributed database load routing, and in particular to a distributed database load dynamic routing method, device, equipment and storage medium. Background Art

[0002] A distributed database usually includes multiple stateless computing nodes to form a cluster. These nodes provide services to the outside world at the same time under the role of load balancing.

[0003] Since there are usually many computing nodes in a distributed database, the probability of abnormalities in computing nodes is relatively high. The computing nodes with abnormalities will rejoin the cluster after returning to normal. The load balancing technology in the prior art usually adopts polling or random algorithms. Polling is to sort each computing node and route the new load to be processed to the computing node in order for processing; the random algorithm is to randomly assign a computing node to process the new load to be processed. Since the above allocation method does not take into account the load conditions of each computing node, the computing nodes that rejoin the cluster after the abnormality often have a relatively low load, while the load of some computing nodes in the cluster is relatively high, resulting in local hot spots. This not only causes a waste of computing resources of the computing nodes, but also further increases the probability of abnormalities of some computing nodes, especially high-load computing nodes. Therefore, there is an urgent need for a distributed database load dynamic routing method, device, equipment and storage medium to solve the above-mentioned problems. Summary of the invention

[0004] The present application provides a distributed database load dynamic routing method, apparatus, device and storage medium to improve the query performance of the database.

[0005] In a first aspect, the present application provides a distributed database load dynamic routing method, the method comprising:

[0006] Get information about the load to be processed and the status of all computing nodes in the cluster;

[0007] Based on the status information of each of the computing nodes, calculate and obtain current load level information of each of the computing nodes;

[0008] Inputting the information of the load to be processed into a preset load resource model, and obtaining the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model;

[0009] Based on the current load level information of each of the computing nodes and the predicted resource occupancy of processing the load to be processed within the future preset time period, the predicted load level of each of the computing nodes when processing the load to be processed within the future preset time period is calculated, and the computing node with the lowest predicted load level is selected as the target computing node, and the load to be processed is routed to the target computing node for processing.

[0010] In a second aspect, the present application provides a distributed database load dynamic routing device, the device comprising:

[0011] The acquisition module is used to obtain information about the load to be processed and the status information of all computing nodes in the cluster;

[0012] A calculation module, configured to calculate current load level information of each of the computing nodes based on the status information of each of the computing nodes;

[0013] A prediction module, used for inputting the information of the load to be processed into a preset load resource model, and obtaining the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model;

[0014] A routing module is used to calculate the predicted load level of each computing node when processing the load to be processed within the future preset time period based on the current load level information of each computing node and the predicted resource occupancy of processing the load to be processed within the future preset time period, select the computing node with the lowest predicted load level as the target computing node, and route the load to be processed to the target computing node for processing.

[0015] In a third aspect, the present application provides an electronic device comprising: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; and at least one memory connected to the at least one bus, wherein the processor is configured to execute a distributed database load dynamic routing method described in the first aspect above.

[0016] In a fourth aspect, the present application further provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the distributed database load dynamic routing method described in the first aspect of the present application.

[0017] The above-mentioned technical scheme provided by the embodiment of the present application has the following advantages compared with the prior art: the distributed database load dynamic routing method provided by the embodiment of the present application obtains the information of the load to be processed and the status information of all computing nodes in the cluster, obtains the current load level information of each computing node according to the status information of the computing node, and then obtains the predicted resource occupancy of each computing node for processing the load to be processed by inputting the information of the load to be processed into the load resource model; finally, according to the current load level information of each computing node and the predicted resource occupancy for processing the load to be processed, selects the computing node with the lowest predicted load level when processing the load to be processed, and uses it as the routing object of the load to be processed so that it can process the load to be processed. Compared with the load balancing technology in the prior art that uses polling or random algorithms, which causes the problem of unbalanced load distribution and the problem of waste of computing node resources and increased probability of abnormalities caused by load imbalance, the technical solution adopted in the present application can adaptively and dynamically route the load to be processed to the computing node with the lowest resource occupancy when processing the load, thereby effectively balancing the load of each computing node in the cluster, avoiding the generation of local hotspots to the greatest extent, reducing the probability of computing node abnormalities caused by load imbalance, and achieving full utilization of the computing resources of the computing nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0020] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0021] Figure 1 A schematic diagram of a process flow of a distributed database load dynamic routing method provided by an embodiment of the present application;

[0022] Figure 2 A schematic diagram of the structure of a distributed database load dynamic routing device provided in an embodiment of the present application;

[0023] Figure 3A schematic diagram of the workflow of a distributed database load dynamic routing device provided in an embodiment of the present application;

[0024] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings 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 making creative work are within the scope of protection of this application.

[0026] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0027] In order to solve the problem in the prior art that local hot spots are caused by unbalanced load on computing nodes, which in turn causes a waste of computing resources of computing nodes and increases the probability of abnormalities in some computing nodes, the present application provides a distributed database load dynamic routing method that can effectively balance the load of each computing node in the cluster, avoid the generation of local hot spots to the greatest extent, and reduce the probability of computing node abnormalities caused by load imbalance.

[0028] Figure 1 A distributed database load dynamic routing method provided in an embodiment of the present application, referring to Figure 1 , a distributed database load dynamic routing method provided by the present application includes:

[0029] S1: Obtain information about the load to be processed and the status information of all computing nodes in the cluster.

[0030] The information of the load to be processed is obtained from the business end, which is the initiator of the business application and the transmission starting point of the business traffic.

[0031] In a feasible embodiment of the present application, the status information of each computing node can be obtained from the cluster according to a fixed time period. It should be understood that the present application does not limit the specific time period for obtaining the status information of the computing node, which can be set according to actual conditions. Preferably, the present application adopts a time period of 5 seconds, that is, the status information of each computing node is obtained from the cluster every 5 seconds.

[0032] In a feasible embodiment of the present application, the status information of the computing node includes but is not limited to the unique information identifier of the connected load corresponding to the current computing node, the connection duration of each connected load, the resource occupancy of each connected load, and the overall resource occupancy of the current computing node. Among them, the unique information identifier of the load is used to identify the load to determine the load that the computing node is currently processing or about to process. In a feasible embodiment of the present application, the unique information identifier of the load includes but is not limited to the source load IP, business module name and load connection sequence number of the service end corresponding to the load. Among them, the service end is the service application initiator, which is the transmission starting point of the service traffic, and the source load IP is an IP address routed to the service end for identifying the source of the service end request traffic. The service module name refers to the identifier of different services or functions in the database, such as "query service", "data storage service", etc.; the load connection sequence number is used to distinguish different database load connections. For example, in the unique information identifier of a load, the source load IP is 192.168.1.10, the service module name is "query service", and the load connection sequence number is 012.

[0033] In a feasible embodiment of the present application, the resource occupancy of each connected load includes but is not limited to the CPU occupancy information, memory occupancy information and I / O occupancy information of each connected load, and the overall resource occupancy of the current computing node includes but is not limited to the CPU occupancy information, memory occupancy information and I / O occupancy information of the current computing node.

[0034] S2: Based on the status information of each computing node, calculate the current load level information of each computing node.

[0035] Specifically, after the status information of each computing node is obtained in step S1, the current load level of each computing node is calculated based on the overall resource occupancy of the current computing node in the status information. For example, one of the items can be selected as the basis for judging the current load level of the computing node based on the connection duration, CPU occupancy information, memory occupancy information, and I / O occupancy information of the connected load of the current computing node, or weights can be assigned to the connection duration, CPU occupancy information, memory occupancy information, and I / O occupancy information of the connected load, and the overall occupancy of the CPU, memory, and I / O is comprehensively calculated as the current load level of the computing node. The specific method can be selected according to the actual situation, and this application does not limit this.

[0036] S3: Input the information of the load to be processed into a preset load resource model, and obtain the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model.

[0037] Specifically, the load resource model is established by obtaining and fitting the current load level information of each computing node in the corresponding life cycle when processing various historical loads. The load resource model stores the predicted resource usage of each computing node in processing various historical loads. For example, assuming that a computing node has processed a load with a source IP of 192.168.1.10, a business module name of "query service", and a load connection number of 012, and the life cycle of the load is 20s, then the load resource model stores the load level curve of the computing node in the 20s life cycle of processing the load. According to the load level curve, the predicted resource usage of the computing node in processing the load at various time points in the 20s life cycle can be obtained.

[0038] After the information of the pending load sent by the business end is input into the load resource model, the model will match the corresponding historical load according to the information of the pending load, that is, match the corresponding historical load according to the source load IP, business module name and load connection sequence number of the pending load, and then use the predicted resource occupancy of the historical load when it is processed by each computing node during its life cycle as the predicted resource occupancy of the pending load when it is processed by each computing node.

[0039] It should be understood that the above steps S2 and S3 may be performed simultaneously or separately, and the present application does not limit the specific execution order.

[0040] S4: Based on the current load level information of each computing node and the predicted resource occupancy of processing the load to be processed within a future preset time period, calculate the predicted load level of each computing node when processing the load to be processed within a future preset time period, select the computing node with the lowest predicted load level as the target computing node, and route the load to be processed to the target computing node for processing.

[0041] Specifically, in step S2, the current load level information of each computing node has been learned, and in step S3, the predicted resource occupancy of each computing node in processing the load to be processed in the future preset time period has been learned. Therefore, these two parameters can be used to calculate the predicted load level of each computing node in processing the load to be processed in the future preset time. The specific steps are as follows:

[0042] S401: Using the current load level information of each computing node as the basic load.

[0043] S402: The predicted resource occupancy of each computing node in processing the load to be processed within a future preset time period is used as the increased load; the duration of the future preset time period is the same as the life cycle duration of the load to be processed.

[0044] S403: Add the basic load and the increased load to obtain a predicted load level within a future preset time period.

[0045] For example, suppose there are three computing nodes in a cluster, among which the current load level of computing node No. 1 (i.e., basic load) is 15%, the current load level of computing node No. 2 is 10%, and the current load level of computing node No. 3 is 20%. The load resource model stores the predicted resource occupancy of the historical loads of the three computing nodes, whose source IP is 192.168.1.10, whose business module name is "query service", and whose load connection sequence number is 012. The predicted resource occupancy (i.e., increased load) of node No. 1 for processing the load is 10%, the predicted resource occupancy of node No. 2 for processing the load is 30%, and the predicted resource occupancy of node No. 3 for processing the load is 15%. Then the predicted load level of computing node No. 1 for processing the load is 15%+10%=25%, the predicted load level of computing node No. 2 for processing the load is 10%+30%=40%, and the predicted load level of computing node No. 3 for processing the load is 20%+15%=35%. After respectively calculating the predicted load levels of the three computing nodes when processing the load to be processed in the future preset time period, the computing node with the lowest predicted load level, that is, computing node No. 1, is selected as the target computing node, and the load to be processed is routed to computing node No. 1 for processing. This load routing method avoids the situation where the load level of one or some computing nodes in the cluster is high while the load level of other computing nodes is low, and can fully utilize the computing resources of the computing nodes, while reducing the probability of computing node abnormality due to high load of local computing nodes.

[0046] For the predicted resource occupancy of each computing node processing the load stored in the load resource model, if only static data is maintained without updating it, it will inevitably lead to an increasing deviation between the predicted resource occupancy and the actual resource occupancy of the computing node processing the load, thereby reducing the accuracy of the prediction. Therefore, in a feasible embodiment of the present application, in order to achieve closed-loop feedback of data in the load resource model and improve the accuracy of the predicted resource occupancy output by the load resource model, after routing the load to be processed to the target computing node for processing in step S4, it also includes: obtaining the actual resource occupancy of the target computing node during the entire life cycle corresponding to the load to be processed, and adjusting the load resource model based on the actual resource occupancy. Specifically, adjusting the load resource model based on the actual resource occupancy includes the following steps:

[0047] S411: Obtain a unique information identifier of the load to be processed.

[0048] S412: Searching for a historical load matching the unique information identifier in the load resource model.

[0049] S413: Based on the actual resource occupancy, the predicted resource occupancy of the target computing node in the load resource model for processing the historical load is updated.

[0050] Taking the above three computing nodes as an example, it is assumed that the load resource model stores the predicted resource occupancy of the historical loads whose source load IP is 192.168.1.10, the business module name is "query service", and the load connection number is 012. The source load IP, business module name and load connection number of the load to be processed initiated by the business end all match the historical load. After executing the above step S4, i.e. selecting computing node No. 1 with the lowest predicted load level as the target computing node and routing the load to be processed to computing node No. 1 for processing, it is necessary to obtain the actual resource occupancy of computing node No. 1 occupied by the load to be processed during its life cycle when computing node No. 1 processes the load to be processed, and then search the load resource model for the historical load that matches the load to be processed according to the unique information identifier of the load to be processed, and update the predicted resource occupancy corresponding to the execution of the historical load by computing node No. 1 in the load resource model according to the actual resource occupancy of the load to be processed, and at the same time, update the life cycle corresponding to the execution of the historical load by computing node No. 1 in the load resource model according to the actual life cycle of processing the load to be processed. By updating the predicted resource occupancy in the load resource model according to the actual resource occupancy in the actual processing process of the load, a closed-loop feedback of the data in the load resource model is achieved, which can make the predicted resource occupancy in the load resource model closer to the actual resource occupancy, thereby improving the accuracy of the predicted resource occupancy output by the load resource model.

[0051] The technical solution provided by the present application obtains information about the load to be processed and status information of all computing nodes in the cluster, obtains current load level information of each computing node based on the status information of the computing nodes, and then obtains the predicted resource occupancy of each computing node for processing the load to be processed by inputting the information of the load to be processed into a load resource model. Based on the current load level information of each computing node and the predicted resource occupancy for processing the load to be processed, a target computing node with the lowest predicted load level when processing the load to be processed is selected and used as the routing object of the load to be processed so that it can process the load to be processed. Compared with the load balancing technology used in the prior art, which causes problems of unbalanced load distribution, waste of computing node resources and increased probability of abnormalities due to load imbalance, and unbalanced load on computing nodes due to the difference between short connections and long connections in the load of a long-running cluster system, the technical solution adopted in the present application can adaptively and dynamically route the load to be processed to the computing node with the lowest resource occupancy when processing the load, thereby effectively balancing the load of each computing node in the cluster, avoiding the generation of local hot spots to the greatest extent, reducing the probability of computing node abnormalities caused by load imbalance, and fully utilizing the computing resources of the computing nodes. In addition, in the solution of the present application, the load resource model is updated according to the actual resource occupancy of the load to be processed during the process of the target computing node processing the load to be processed, thereby realizing closed-loop feedback of data in the load resource model, and can make the predicted resource occupancy in the load resource model closer to the actual resource occupancy, thereby improving the accuracy of the predicted resource occupancy output by the load resource model, and making the dynamic routing of the load more reasonable.

[0052] Corresponding to the above method embodiment, the present application also provides a distributed database load dynamic routing device, referring to Figure 2 A distributed database load dynamic routing device provided in the present application includes: a collection module 201, a calculation module 202, a prediction module 203 and a routing module 204.

[0053] The acquisition module 201 is used to obtain information about the load to be processed and status information of all computing nodes in the cluster.

[0054] The calculation module 202 is used to calculate the current load level information of each computing node based on the status information of each computing node.

[0055] The prediction module 203 is used to input the information of the load to be processed into a preset load resource model, and obtain the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model.

[0056] The routing module 204 is used to calculate the predicted load level of each computing node when processing the load to be processed within a future preset time period based on the current load level information of each computing node and the predicted resource occupancy of processing the load to be processed within a future preset time period, select the computing node with the lowest predicted load level as the target computing node, and route the load to be processed to the target computing node for processing.

[0057] Specifically, Figure 3 As shown, a distributed database load dynamic routing device provided by the present application obtains the information of the load to be processed sent by the business end through the acquisition module, and obtains the status information of each computing node in the cluster, and then the acquisition module sends the status information of each computing node to the calculation module; the calculation module calculates the current load level information of each computing node based on the status information of each computing node; the prediction module obtains the information of the load to be processed and inputs the preset load resource model, and obtains the predicted resource occupancy of each computing node in the future preset time period output by the load resource model for processing the load to be processed; finally, the routing module calculates the predicted load level of each computing node when processing the load to be processed in the future preset time period based on the current load level information of each computing node and the predicted resource occupancy of processing the load to be processed in the future preset time period, selects the computing node with the lowest predicted load level as the target computing node, and dynamically routes the load to be processed to the target computing node for processing. The target computing node parses and optimizes the load to be processed and then routes it to the corresponding data node, then receives and processes the result set returned by the data node, and summarizes the results after the processing is completed and returns them to the business end. It should be understood that the work of the calculation module and the prediction module can be performed simultaneously or separately, and this application does not limit this.

[0058] like Figure 4 As shown, an embodiment of the present application provides an electronic device, including a processor 411, a communication interface 412, a memory 413 and a communication bus 414, wherein the processor 411, the communication interface 412, and the memory 413 communicate with each other through the communication bus 414, and the memory 413 is used to store computer programs.

[0059] In one embodiment of the present application, the processor 411 is used to implement the distributed database load dynamic routing method provided by any one of the aforementioned method embodiments when executing the program stored in the memory 413, including:

[0060] Get information about the load to be processed and the status of all computing nodes in the cluster;

[0061] Based on the status information of each of the computing nodes, calculate and obtain current load level information of each of the computing nodes;

[0062] Inputting the information of the load to be processed into a preset load resource model, and obtaining the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model;

[0063] Based on the current load level information of each of the computing nodes and the predicted resource occupancy of processing the load to be processed within the future preset time period, the predicted load level of each of the computing nodes when processing the load to be processed within the future preset time period is calculated, and the computing node with the lowest predicted load level is selected as the target computing node, and the load to be processed is routed to the target computing node for processing.

[0064] An embodiment of the present application also 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 the distributed database load dynamic routing method provided in any of the aforementioned method embodiments are implemented.

[0065] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0066] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0067] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0068] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A distributed database load dynamic routing method, characterized in that: include: Get information about the load to be processed and the status of all computing nodes in the cluster; Based on the status information of each of the computing nodes, calculate and obtain current load level information of each of the computing nodes; Inputting the information of the load to be processed into a preset load resource model, and obtaining the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model; Based on the current load level information of each of the computing nodes and the predicted resource occupancy of processing the load to be processed within the future preset time period, the predicted load level of each of the computing nodes when processing the load to be processed within the future preset time period is calculated, and the computing node with the lowest predicted load level is selected as the target computing node, and the load to be processed is routed to the target computing node for processing.

2. The distributed database load dynamic routing method according to claim 1, characterized in that: The load resource model is established by acquiring and fitting the current load level information of each computing node in its corresponding life cycle when processing various historical loads; the load resource model stores the predicted resource occupancy of each computing node in processing various historical loads.

3. The distributed database load dynamic routing method according to claim 2, characterized in that: After routing the load to be processed to the target computing node for processing, the method further includes: The actual resource occupancy of the load to be processed during the entire life cycle of the load to be processed is obtained when the target computing node processes the load to be processed, and the load resource model is adjusted based on the actual resource occupancy.

4. The distributed database load dynamic routing method according to claim 3, characterized in that: Adjusting the load resource model based on the actual resource occupancy includes: Obtaining a unique information identifier of the load to be processed; Searching the load resource model for the historical load that matches the unique information identifier; Based on the actual resource occupancy, the predicted resource occupancy of the target computing node in the load resource model for processing the historical load is updated.

5. The distributed database load dynamic routing method according to claim 1, characterized in that: The status information includes a unique information identifier of a connected load corresponding to the computing node, a connection duration of the connected load, a resource occupancy amount of the connected load, and a total resource occupancy amount of the computing node.

6. The distributed database load dynamic routing method according to claim 5, characterized in that: The unique information identifier includes the source end load IP, the service module name and the load connection sequence number of the service end corresponding to the load.

7. The distributed database load dynamic routing method according to any one of claims 1 to 6, characterized in that: Based on the current load level information of each computing node and the predicted resource occupation amount of processing the load to be processed in the future preset time period, the predicted load level of each computing node when processing the load to be processed in the future preset time period is calculated, including: Using the current load level information of each computing node as the basic load; The predicted resource occupation amount of each computing node in processing the load to be processed within the future preset time period is used as the increased load amount; the duration of the future preset time period is the same as the life cycle duration of the load to be processed; The basic load amount and the increased load amount are added to obtain the predicted load level within the future preset time period.

8. A distributed database load dynamic routing device, characterized in that: The device comprises: The acquisition module is used to obtain information about the load to be processed and the status information of all computing nodes in the cluster; A calculation module, configured to calculate current load level information of each of the computing nodes based on the status information of each of the computing nodes; A prediction module, used for inputting the information of the load to be processed into a preset load resource model, and obtaining the predicted resource occupancy of each computing node for processing the load to be processed within a future preset time period output by the load resource model; A routing module is used to calculate the predicted load level of each computing node when processing the load to be processed within the future preset time period based on the current load level information of each computing node and the predicted resource occupancy of processing the load to be processed within the future preset time period, select the computing node with the lowest predicted load level as the target computing node, and route the load to be processed to the target computing node for processing.

9. An electronic device, characterized in that: include: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor coupled to the at least one bus; At least one memory connected to the at least one bus, wherein the processor is configured to implement the distributed database load dynamic routing method according to any one of claims 1-7.

10. A computer storage medium, characterized in that: Computer executable instructions are stored, and the computer executable instructions are used to execute the distributed database load dynamic routing method according to any one of claims 1 to 7.

Images