Task scheduling method and device for financial database, electronic equipment and storage medium

Abstract

The invention discloses a task scheduling method and device for a financial database, electronic equipment and a storage medium, and relates to the technical field of big data and distributed computing or other related fields, the method comprises the steps of receiving a task request queue of the financial database, and obtaining a computing resource use state of a financial system, the task request queue comprises N database tasks to be executed, and N is a positive integer; calculating a priority score of each database task in the task request queue; allocating computing resources to each database task based on the priority score and the computing resource use state to obtain a resource allocation result; and performing resource scheduling on the task request queue based on the resource allocation result. Through the method and the device, the technical problem of low task resource scheduling efficiency of a financial database in related technologies is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data and distributed computing or other related fields, in particular, to a financial database task scheduling method and device, an electronic device and a storage medium. BACKGROUND

[0002] In the field of task scheduling of financial databases, the existing technology mainly adopts a static scheduling strategy based on first-come-first-served or simple resource matching, which usually only allocates according to the order of task arrival or basic resource demand, and lacks consideration of key factors such as task business value and timeliness requirements.

[0003] In particular, in a high-concurrency scenario, low-value tasks may preempt critical computing resources, causing high-priority tasks such as transaction settlement and risk control to be delayed. At the same time, traditional scheduling systems are difficult to respond to dynamic changes in financial business scenarios in a timely manner, and cannot adjust resource allocation strategies according to real-time load conditions, resulting in the dual problems of low system resource utilization and unstable quality of service for critical business.

[0004] With the continuous expansion of financial business scale, the defects of the existing scheduling method in terms of resource allocation efficiency are increasingly prominent. On the one hand, the fixed priority task processing mechanism is difficult to adapt to the characteristics of real-time fluctuations in the financial market, and is prone to cause task accumulation during periods of dramatic market fluctuations; on the other hand, the lack of intelligent resource dynamic allocation capability makes the system perform poorly when dealing with sudden business peaks, which not only fails to guarantee the real-time requirements of core business, but also makes it difficult to achieve global optimal configuration of computing resources. The above-mentioned low scheduling efficiency directly affects the overall performance of the financial database system.

[0005] In view of the above problems, no effective solutions have been proposed so far. SUMMARY

[0006] The main purpose of the present application is to provide a financial database task scheduling method and device, an electronic device and a storage medium, to at least solve the technical problem of low efficiency of task resource scheduling of financial databases in related technologies.

[0007] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a financial database task scheduling method is provided, which comprises: receiving a task request queue of a financial database and obtaining a computing resource usage state of a financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer; calculating the priority score of each database task in the task request queue; allocating computing resources for each database task based on the priority score and the computing resource usage state to obtain a resource allocation result; and performing resource scheduling on the task request queue based on the resource allocation result.

[0008] Further, the step of acquiring the computing resource usage state of the financial system comprises: collecting state parameters of M computing nodes in the financial system, wherein the state parameters comprise CPU usage, memory occupancy, and network bandwidth occupancy, and M is a positive integer; performing normalization processing on the collected state parameters to obtain resource index data in a standard format; performing weighted calculation on the resource index data of each computing node according to a first preset weight configuration to obtain a load evaluation value of each computing node; and generating a system-level resource usage state based on the load evaluation values of all computing nodes.

[0009] Further, the step of calculating the priority score of each database task in the task request queue comprises: for each database task, acquiring task attribute parameters of the database task, wherein the task attribute parameters comprise an influence range parameter, a time limit requirement parameter, and a resource demand parameter; and performing weighted calculation on the influence range parameter, the time limit requirement parameter, and the resource demand parameter according to a second preset weight configuration to obtain the priority score of each database task.

[0010] Further, the step of allocating computing resources to each database task based on the priority score and the computing resource usage state to obtain a resource allocation result comprises: classifying each database task in the task request queue based on the priority score to obtain a task level mark; classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification mark; and performing resource allocation based on the task level mark and the node classification mark to obtain the resource allocation result.

[0011] Further, the step of classifying each database task in the task request queue based on the priority score to obtain a task level mark comprises: for each database task, matching the priority score with a preset task classification threshold to obtain a first matching result, wherein the preset task classification threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; in a case where the first matching result indicates that the priority score is greater than or equal to the first threshold, marking the database task as a high-priority task; in a case where the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold, marking the database task as a medium-priority task; and in a case where the first matching result indicates that the priority score is less than the second threshold, marking the database task as a low-priority task.

[0012] Further, the step of classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification label comprises: for each computing node, matching the load evaluation value of the computing node indicated by the computing resource usage state with a preset node classification threshold to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; in the case that the second matching result indicates that the load evaluation value is greater than or equal to the third threshold, marking the computing node as an overload node; in the case that the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold, marking the computing node as an available node; in the case that the second matching result indicates that the load evaluation value is less than the fourth threshold, marking the computing node as an idle node.

[0013] Further, the step of performing resource allocation based on the task level label and the node classification label to obtain a resource allocation result comprises: configuring a resource exclusive allocation strategy for high-priority tasks, wherein the resource exclusive allocation strategy means that the high-priority tasks are preferentially allocated to idle nodes, in the case that the idle nodes are fully occupied, the remaining high-priority tasks that are not allocated are allocated to available nodes, and the high-priority tasks are prohibited from being allocated to overload nodes; configuring a resource elastic allocation strategy for medium-priority tasks, wherein the resource elastic allocation strategy means that after the high-priority task allocation is completed, in the case that there are idle nodes that are not allocated, the medium-priority tasks are preferentially allocated to the idle nodes, the medium-priority tasks that are not allocated are allocated to the available nodes, and the medium-priority tasks are prohibited from being allocated to the overload nodes; configuring a resource limited allocation strategy for low-priority tasks, wherein the resource limited allocation strategy means that after the high-priority tasks and the medium-priority tasks are all allocated, in the case that there are idle nodes that are not allocated, the low-priority tasks are preferentially allocated to the idle nodes, the low-priority tasks that are not allocated are allocated to the available nodes, and the low-priority tasks are prohibited from being allocated to the overload nodes; performing resource allocation operations based on the resource exclusive allocation strategy, the resource elastic allocation strategy, and the resource limited allocation strategy to obtain the resource allocation result.

[0014] To achieve the above object, according to another aspect of the present application, there is further provided a task scheduling device of a financial database, comprising: a receiving unit configured to receive a task request queue of the financial database and acquire a computing resource usage state of a financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer; a computing unit configured to calculate a priority score of each database task in the task request queue; an allocating unit configured to allocate computing resources for each database task based on the priority score and the computing resource usage state to obtain a resource allocation result; and an executing unit configured to perform resource scheduling on the task request queue based on the resource allocation result.

[0015] Further, the receiving unit comprises: a collecting module configured to collect state parameters of M computing nodes in the financial system, wherein the state parameters comprise CPU usage, memory occupancy and network bandwidth occupancy, and M is a positive integer; a normalization processing module configured to perform normalization processing on the collected state parameters to obtain resource index data in a standard format; a first weighting calculation module configured to perform weighting calculation on the resource index data of each computing node according to a first preset weight configuration to obtain a load evaluation value of each computing node; and a generating module configured to generate the resource usage state at the system level based on the load evaluation values of all the computing nodes.

[0016] Further, the computing unit comprises: an acquiring module configured to acquire, for each database task, a task attribute parameter of the database task, wherein the task attribute parameter comprises an influence range parameter, a time limit requirement parameter and a resource demand parameter; and a second weighting calculation module configured to perform weighting calculation on the influence range parameter, the time limit requirement parameter and the resource demand parameter according to a second preset weight configuration to obtain the priority score of each database task.

[0017] Further, the allocating unit comprises: a grading module configured to grade each database task in the task request queue based on the priority score to obtain a task level mark; a classification module configured to classify each computing node in the financial system based on the computing resource usage state to obtain a node classification mark; and an allocating module configured to perform resource allocation based on the task level mark and the node classification mark to obtain the resource allocation result.

[0018] Further, the grading module comprises: a first matching submodule, configured to match, for each of the database tasks, the priority score with a preset task grading threshold to obtain a first matching result, wherein the preset task grading threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; a first marking submodule, configured to mark the database task as a high-priority task if the first matching result indicates that the priority score is greater than or equal to the first threshold; a second marking submodule, configured to mark the database task as a medium-priority task if the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold; and a third marking submodule, configured to mark the database task as a low-priority task if the first matching result indicates that the priority score is less than the second threshold.

[0019] Further, the classification module comprises: a second assignment submodule, configured to match, for each of the computing nodes, a load evaluation value of the computing node indicated by the computing resource usage state with a preset node classification threshold to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; a fourth marking submodule, configured to mark the computing node as an overloaded node if the second matching result indicates that the load evaluation value is greater than or equal to the third threshold; a fifth marking submodule, configured to mark the computing node as an available node if the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold; and a sixth marking submodule, configured to mark the computing node as an idle node if the second matching result indicates that the load evaluation value is less than the fourth threshold.

[0020] Further, the allocation module comprises: a first configuration submodule, configured to configure a resource exclusive allocation strategy for high-priority tasks, wherein the resource exclusive allocation strategy refers to preferentially allocating the high-priority tasks to idle nodes, in the case that the idle nodes are fully occupied, preferentially allocating the remaining high-priority tasks which are not allocated to available nodes, and prohibiting the high-priority tasks from being allocated to overloaded nodes; a second configuration submodule, configured to configure a resource elastic allocation strategy for medium-priority tasks, wherein the resource elastic allocation strategy refers to, after the high-priority tasks are allocated, in the case that there are idle nodes which are not allocated, preferentially allocating the medium-priority tasks to the idle nodes, preferentially allocating the medium-priority tasks which are not allocated to the available nodes, and prohibiting the medium-priority tasks from being allocated to the overloaded nodes; a third configuration submodule, configured to configure a resource limited allocation strategy for low-priority tasks, wherein the resource limited allocation strategy refers to, after the high-priority tasks and the medium-priority tasks are all allocated, in the case that there are idle nodes which are not allocated, preferentially allocating the low-priority tasks to the idle nodes, preferentially allocating the low-priority tasks which are not allocated to the available nodes, and prohibiting the low-priority tasks from being allocated to the overloaded nodes; and an execution submodule, configured to execute resource allocation operations based on the resource exclusive allocation strategy, the resource elastic allocation strategy and the resource limited allocation strategy, to obtain the resource allocation result.

[0021] To achieve the above object, according to another aspect of the present application, there is further provided a computer readable storage medium comprising a stored computer program, wherein the computer readable storage medium controls a device in which the computer readable storage medium is located to perform the task scheduling method of the financial database according to any one of the above aspects when the computer program is executed.

[0022] To achieve the above object, according to another aspect of the present application, there is further provided an electronic device comprising one or more processors and a memory, the memory being configured to store one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the task scheduling method of the financial database according to any one of the above aspects.

[0023] To achieve the above object, according to another aspect of the present application, there is further provided a computer program product comprising computer instructions, wherein the computer instructions, when executed by a processor, implement the steps of the task scheduling method of the financial database according to any one of the above aspects.

[0024] The application provides a financial database task scheduling method, which comprises the following steps: receiving a task request queue of a financial database, and obtaining a computing resource usage state of a financial system, wherein the task request queue comprises N database tasks to be executed, N is a positive integer, the priority score of each database task in the task request queue is calculated, then the computing resource of each database task is allocated based on the priority score and the computing resource usage state to obtain a resource allocation result, and finally the resource scheduling is performed on the task request queue based on the resource allocation result.

[0025] In the application, the intelligent priority evaluation and dynamic resource allocation are combined, the reinforcement learning algorithm and multi-dimensional task attribute analysis are fused, the computing resource and the task priority requirement are accurately matched, the technical effect of improving the financial database task scheduling efficiency and resource utilization efficiency is realized, and specifically, after the task request queue of the financial database is received, the deep reinforcement learning technology is used to comprehensively analyze each database task to be executed, the priority score of each task is accurately calculated, then the resource allocation strategy is dynamically generated based on the real-time monitored computing resource usage state and the intelligent scheduling algorithm, the high-priority task can quickly obtain sufficient computing resource, and the low-priority task can obtain the opportunity of reasonable scheduling and execution without affecting the overall efficiency of the system, the resource waste and the computing bottleneck are effectively avoided, the response speed and the processing capacity of the system are significantly improved, the continuity and stability of the core business can be ensured during the financial transaction peak, and the technical problem of low efficiency of the task resource scheduling of the financial database in the related art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0026] The drawings constituting a part of the application are used to provide further understanding of the application, the illustrative embodiments of the application and the description thereof are used to explain the application, and do not constitute improper limitation on the application. In the drawings:

[0027] Figure 1 Fig. 1 shows a hardware structure block diagram of a computer terminal (or a mobile device) for implementing the financial database task scheduling method;

[0028] Figure 2 Fig. 2 is a flowchart of an optional financial database task scheduling method according to an embodiment of the application;

[0029] Figure 3 Fig. 3 is a schematic diagram of an optional financial database task scheduling device according to an embodiment of the application;

[0030] Figure 4 Fig. 4 is a structure block diagram of an electronic device for executing the financial database task scheduling method according to an embodiment of the application. DETAILED DESCRIPTION

[0031] In order to make the person skilled in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by the person skilled in the art without creative labor should belong to the protection scope of the present application.

[0032] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0033] In order to facilitate the person skilled in the art to understand the present application, the following explains some terms or nouns involved in each embodiment of the present application:

[0034] MPP database, Massively Parallel Processing Database, a distributed database architecture that uses a large number of independent computing nodes to process data in parallel, each node has its own processor, memory and storage resources, especially suitable for processing large-scale data sets and complex queries. Under the MPP architecture, data and computing tasks can be effectively divided and executed in parallel, thereby significantly improving the processing speed and throughput of the database system.

[0035] Reinforcement Learning, reinforcement learning, a branch of machine learning, through trial-and-error mechanism and reward feedback, so that the intelligent agent can learn how to take the best action in an unknown or dynamic environment to achieve the goal. Reinforcement learning is used in the present application to build a scheduling algorithm that can dynamically adjust the priority of tasks and resource allocation strategy according to historical task execution data and current system state.

[0036] It should be noted that the task scheduling method and device of the financial database in the present application can be used in the field of big data and distributed computing technology in the case of real-time processing and analysis of massive transaction data, and can also be used in any field other than the field of big data and distributed computing technology in the case of real-time processing and analysis of massive transaction data. The application field of the task scheduling method and device of the financial database in the present application is not limited.

[0037] It should be noted that the related information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, storage, processing, transmission, provision, disclosure, use and processing of related data comply with the laws, regulations and standards of the relevant region, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal. For example, the system and the interface between the related users or institutions are provided, before obtaining the related information, the interface needs to send an acquisition request to the aforementioned user or institution, and after receiving the consent information fed back by the aforementioned user or institution, the related information is acquired.

[0038] The information collection (for example, user voice, video, text collection) and analysis operation involved in the present application have provided corresponding operation portal for the user to choose to agree or refuse the automatic decision result when executing; if the user chooses to refuse, the expert decision process is entered.

[0039] The following embodiments of the present application can be applied to various systems / applications / devices that need to perform financial database task scheduling and computing resource optimization, and can realize intelligent dynamic resource allocation products based on task priority. The present application uses reinforcement learning algorithm to analyze the attributes of each database task in the task request queue and calculate the priority score, and then allocates computing resources to each database task based on the priority score and the computing resource usage state, which can better identify and respond to critical business needs, while ensuring the efficiency and fairness of resource allocation.

[0040] The present application also realizes adaptive adjustment of scheduling strategy through real-time resource monitoring and feedback optimization, which can guarantee the stable operation of the system and the optimal utilization of resources even in complex and variable business environment.

[0041] The present application will be described in detail below in conjunction with various embodiments.

[0042] Embodiment one

[0043] According to the embodiment of the present application, an embodiment of a task scheduling method of a financial database is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown.

[0044] The embodiment of the task scheduling method of the financial database provided by the embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing the task scheduling method of the financial database is shown. As shown in the figure, Figure 1 The computer terminal 10 (or mobile device) can include one or more processors 102 (the processor 102 can include but is not limited to a microprocessor MCU or a programmable logic device FPGA processing device), a memory 104 for storing data, and a transmission device 106 for communication function. In addition, it can also include a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which can be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. Those skilled in the art can understand that, Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can also include more or less components than those shown in the figure, or have a different configuration from that shown in the figure. Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can also include more or less components than those shown in the figure, or have a different configuration from that shown in the figure. Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can also include more or less components than those shown in the figure, or have a different configuration from that shown in the figure.

[0045] It should be noted that the one or more processors 102 and / or other data processing circuits described above can be referred to herein as "data processing circuits" in general. The data processing circuit can be embodied in whole or in part as software, hardware, firmware or any combination thereof. In addition, the data processing circuit can be a single independent processing module, or any one of the other elements combined into the computer terminal 10 (or mobile device) in whole or in part. As referred to in the embodiments of the present application, the data processing circuit is a processor control (for example, the selection of the variable resistance terminal path connected with the interface).

[0046] The memory 104 can be used to store software programs of application software and modules, such as program instructions / data storage devices corresponding to the task scheduling method of the financial database in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the task scheduling method of the financial database described above. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include memories remotely arranged with respect to the processor 102, which can be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0047] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to be able to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet in a wireless manner.

[0048] The display can be, for example, a touch screen type liquid crystal display (LCD), which can enable a user to interact with the user interface of the computer terminal 10 (or mobile device).

[0049] Under the above-mentioned operating environment, the present application provides a task scheduling method of a financial database as shown in Figure 2 The embodiment of the present application is a task scheduling method of a financial database, the implementation subject of the method is an MPP distributed database scheduling system, which is combined with artificial intelligence and reinforcement learning technology, and is used in a financial big data real-time processing scene, in particular, solves the problem of low efficiency of task priority identification and dynamic resource allocation, through intelligent task priority evaluation and dynamic resource scheduling algorithm means, including specific steps of task priority model construction, resource state monitoring, intelligent scheduling decision, resource allocation and task execution, and feedback optimization cycle, to achieve the purposes of improving key business response speed, optimizing resource utilization, enhancing scheduling flexibility and system reliability.

[0050] The embodiments of the present application will be described in detail below in combination with various specific steps.

[0051] Figure 2 is a flowchart of an optional task scheduling method of a financial database according to the embodiments of the present application, as shown in Figure 2As shown, the method comprises the following steps:

[0052] In step S201, a task request queue of a financial database is received, and a computing resource usage state of a financial system is obtained, wherein the task request queue contains N database tasks to be executed, and N is a positive integer.

[0053] Specifically, the financial database refers to a database system specially used for storing, managing and processing financial industry data, to support the daily business operation of financial institutions, involving fund clearing, transaction record, risk management, compliance audit and other core fields. In the context of the big data era, the financial database adopts the MPP (Massively Parallel Processing) architecture to quickly process massive transaction data through parallel computing, meeting the real-time needs of financial business.

[0054] The task request queue is a data structure used to temporarily store all pending database tasks in the financial database system. The pending database tasks can include but are not limited to data queries, data analysis, data updates or data loading operations. Due to the characteristics of financial business, the task request queue will receive a large number of task requests from different business systems or users with different urgency and business value at any time.

[0055] The database task refers to any data processing activity initiated in the financial database, which can be reading data, updating records, executing queries, loading new data or performing complex data analysis, etc. Each database task has specific attributes such as required resource type, resource demand, business value and urgency.

[0056] The computing resource usage state refers to the actual resource occupation of each computing node in the financial system, including processor (CPU) utilization, memory occupancy, storage space usage and network bandwidth consumption, etc. Key indicators. Used to reflect the current load level and available computing capacity of the financial system. Dynamic monitoring of the computing resource usage state helps the system to understand the working state of each node in real time, so as to make reasonable scheduling decisions and avoid resource waste and system bottlenecks.

[0057] The above step S201 embodies the preliminary processing capability of the system to receive tasks and the real-time perception capability of the resource state. The execution of step S201 can enable the financial database scheduling system to fully understand the current business demand and system resource status.

[0058] Optionally, in the task scheduling method of the financial database provided by the embodiment of the application, the step of obtaining the computing resource usage state of the financial system comprises: collecting state parameters of M computing nodes in the financial system, wherein the state parameters comprise: CPU usage rate, memory occupancy rate and network bandwidth occupancy rate, M is a positive integer; performing normalization processing on the collected state parameters to obtain resource index data in a standard format; performing weighted calculation on the resource index data of each computing node according to a first preset weight configuration to obtain a load evaluation value of each computing node; and generating a system-level resource usage state based on the load evaluation values of all computing nodes.

[0059] An optional embodiment, in the task scheduling method of the financial database provided by the embodiment of the application, the step of obtaining the computing resource usage state of the financial system comprises four key parts of collecting, normalization processing, load evaluation and system-level resource state generation of the computing node state parameters.

[0060] Specifically, the collection range in the stage of collecting node state parameters covers all computing nodes in the financial system, and these nodes are the constituent parts of the computing resource pool of the MPP database. The state parameters mainly include CPU usage rate, memory occupancy rate and network bandwidth occupancy rate, which are used to reflect the real-time computing capacity and network communication condition of each computing node.

[0061] Normalization processing refers to data standardization processing on the collected state parameters, which converts the resource usage data of different nodes into index values under the same standard, so as to facilitate unified comparison and analysis, for example, scaling the parameter value to 0 to 1 or adjusting according to the preset standard range.

[0062] Load evaluation value calculation refers to that the resource index data of each computing node is weighted calculated by considering the importance difference of different resources in task execution, for example, since high-computing-intensive tasks are more dependent on CPU performance, the CPU usage rate is given a higher weight.

[0063] The load evaluation value of each computing node is obtained by weighted calculation, which comprehensively reflects the current resource occupancy of each computing node and provides a quantitative reference for resource allocation decision. The higher the load evaluation value is, the more saturated the resource utilization degree of the node is, and vice versa.

[0064] Further, based on the load evaluation values of all computing nodes, a system-level resource usage state can be generated, which can overview the resource allocation and usage of the entire MPP database, help the scheduling algorithm to understand the system load more comprehensively, and thus make more optimized resource scheduling decisions.

[0065] The embodiment of the present application can evaluate the feasibility of task execution based on more specific and quantitative information by collecting and processing node state parameters in detail in combination with scheduling algorithms, effectively improve the accuracy of resource allocation, and the generation of load evaluation value and system-level resource state can also enable the scheduling system to quickly identify resource bottlenecks and idle nodes, reduce the waiting time of high-priority tasks, and improve the overall response speed of the system. The effect is remarkable when processing large-scale transaction data.

[0066] Through the implementation of the above steps, the embodiment of the present application not only solves the problems of uneven resource allocation and lack of quantitative basis for scheduling decisions in traditional MPP database scheduling, but also realizes the beneficial effects of optimizing resource allocation strategy, improving system response speed, and enhancing stability and reliability.

[0067] Step S202, calculate the priority score of each database task in the task request queue.

[0068] Specifically, the priority score of each database task in the task request queue is the result of comprehensive evaluation based on multiple attributes of the task, which can specifically include the following task attributes:

[0069] Business value, used to measure the impact of the task on the core business of the enterprise. High business value tasks usually involve key business decisions or directly affect customer service, and therefore should be executed in priority. For example, transaction settlement, risk control, compliance check, etc. are considered as high business value tasks in the financial field.

[0070] Urgency, used to evaluate the time sensitivity of the task. For example, tasks with high urgency may come from real-time transaction processing, emergency report generation, or emergency response scenarios, and need to be completed within a limited time to avoid business interruption or loss.

[0071] Resource demand, which can be evaluated according to the amount and type of computing resources (such as CPU, memory, storage) required by the task. Tasks with large resource demand and special types may cause great pressure on the system, affecting the normal execution of other tasks.

[0072] By calculating the priority score, tasks can be classified into three priority levels: high, medium and low, so that different levels of tasks are given corresponding priority in resource allocation, ensuring that important and urgent tasks are processed in time, while optimizing resource utilization, avoiding low-priority tasks from occupying too much computing resources, and affecting the overall system efficiency and response time.

[0073] In some optional embodiments, the process of calculating the priority score should also take into account the specific context and business environment of the task, such as the importance of certain tasks significantly increases in a certain period of time (such as the generation of monthly financial statements), or the resource demand of certain tasks temporarily increases due to external factors. Incorporating these dynamic factors into the priority model can ensure the flexibility and adaptability of the scheduling strategy.

[0074] Optionally, in the task scheduling method of the financial database provided by the embodiment of the application, the step of calculating the priority score of each database task in the task request queue comprises: for each database task, obtaining the task attribute parameters of the database task, wherein the task attribute parameters comprise: influence range parameters, time limit requirement parameters and resource demand parameters; and performing weighted calculation on the influence range parameters, the time limit requirement parameters and the resource demand parameters according to a second preset weight configuration to obtain the priority score of each database task.

[0075] It should be noted that the calculation of the priority score of each database task in the task request queue is an intelligent scheduling strategy that deeply analyzes the task attribute parameters and determines the priority of the task by using weighted calculation, thereby ensuring that the resource allocation meets the actual business needs.

[0076] Specifically, the influence range parameters are used to measure the influence degree of the task on the overall business system, including the amount of data involved, the number of users affected, and the influence on business continuity and security. Tasks with high influence range need to be processed in priority.

[0077] The time limit requirement parameters are used to reflect the time limit that the task must be completed, including the hard deadline and the soft expected time. The hard deadline is more stringent, and exceeding the time limit will have serious consequences; while the soft expected time is more used to reflect the expectation of the task completion speed.

[0078] The resource demand parameters are used to evaluate the computing resources required for task execution, including CPU cycles, memory space, storage requirements and network transmission volume. Generally, tasks with high resource demand have greater load pressure on the system, and the execution time and resource allocation need to be carefully decided.

[0079] Further, the embodiment of the application performs weighted calculation on the influence range parameters, the time limit requirement parameters and the resource demand parameters by using a second preset weight configuration, wherein the weight reflects the relative importance of each attribute parameter in a specific scenario. For example, when processing urgent transactions, the weight of the time limit requirement parameters will significantly increase.

[0080] In the embodiment of the present application, the priority score of each task can be accurately evaluated through detailed analysis and weighted calculation of the task attribute parameters, so that more reasonable scheduling decisions can be made in the case of limited resources, high-value and high-urgency tasks are preferentially processed, and resource waste and low-efficiency operations are avoided.

[0081] The weight configuration can be adjusted according to actual business needs to flexibly respond to various business scenarios, whether it is regular daily operation or emergency handling, and a quick response can be made to dynamically adjust the task execution order, thereby improving the overall response capability and adaptability of the system.

[0082] By converting the task attribute parameters into priority scores, not only can the execution order of the tasks be clearly distinguished, but also the computing resources can be reasonably allocated based on the actual needs of each task, avoiding the blindness and inefficiency in resource allocation, and significantly improving the resource utilization efficiency.

[0083] In summary, through the steps of fine task attribute parameter analysis and weighted priority score calculation, the embodiment of the present application not only solves the problems of unclear task priority and unfair resource allocation in traditional scheduling methods, but also achieves multiple beneficial effects such as business process optimization, user satisfaction improvement, and operation cost reduction.

[0084] Step S203, based on the priority score and the computing resource usage state, each database task is allocated computing resources to obtain a resource allocation result. Step S203 involves real-time monitoring of the existing resource state and dynamic scheduling decisions based on task priority.

[0085] Specifically, the resource usage of all nodes under the MPP distributed database architecture needs to be continuously monitored, including CPU utilization, memory occupation, disk I / O speed, network bandwidth and other indicators. Then, based on the task priority score, a dynamic resource allocation strategy is adopted, giving priority to high-priority tasks and allocating more computing resources to ensure that critical tasks can be quickly responded to, meeting the timeliness and performance requirements. At the same time, for low-priority tasks, a delay scheduling or resource preemption strategy is adopted, that is, the task is executed when resources allow, or when a high-priority task arrives, the execution is moderately interrupted to release resources to higher-priority tasks.

[0086] The resource allocation algorithm in the embodiment of the present application can use an intelligent scheduling algorithm, such as a reinforcement learning algorithm, to optimize the resource allocation process. The algorithm learns from historical task execution data and current resource state, and then is applied to dynamically adjust the resource allocation strategy to achieve global optimization. For example, the algorithm learns that the resource utilization of certain nodes is low at a certain time point, and these nodes are given priority in resource allocation to improve the overall utilization efficiency of resources.

[0087] In addition, the resource allocation result is fed back to the task execution module in real time to ensure that the task can be executed according to the predetermined strategy. In addition, the actual resource consumption and the task execution state in the execution process are recorded for subsequent feedback optimization cycles to further improve the accuracy of the scheduling strategy.

[0088] In another optional embodiment, the resource allocation process can also be configured to adaptively adjust the strategy according to the real-time changes in the system state and the task priority. For example, in the case of a sudden emergence of a large number of high-priority task requests, the resource allocation is quickly re-evaluated to preferentially meet the needs of high-priority tasks, and other tasks can be arranged in the case of sufficient computing resources for high-priority tasks.

[0089] The resource preemption and delay scheduling mechanism are also supported in the embodiments of the present application. In the case of limited resources, resource preemption allows high-priority tasks to interrupt low-priority tasks if necessary, and delay scheduling can execute non-urgent tasks when resources are relatively idle, reducing resource competition and improving overall efficiency.

[0090] Through the above resource allocation process, the present application realizes the optimal utilization of resources of the MPP distributed database system, ensures that all tasks, especially high-priority tasks, can be executed at the most appropriate time and with the most sufficient resources, thereby improving the response speed, processing capacity and overall performance of the system, while reducing resource waste and operation and maintenance costs.

[0091] Optionally, in the task scheduling method for the financial database provided by the embodiments of the present application, the step of allocating computing resources to each database task based on the priority score and the computing resource usage state to obtain a resource allocation result comprises: classifying each database task in the task request queue based on the priority score to obtain a task level mark; classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification mark; and performing resource allocation based on the task level mark and the node classification mark to obtain the resource allocation result.

[0092] It should be noted that the step of allocating computing resources to each database task based on the priority score and the computing resource usage state is a key step to ensure efficient resource utilization, response speed and service quality.

[0093] First, based on the priority score calculated previously, each database task in the task request queue is ranked, and the tasks are divided into three levels, high, medium and low, represented by task level markers. High priority tasks have the highest priority in resource allocation to ensure rapid start and completion; compared with high priority tasks, medium priority tasks have a certain time limit but are not as urgent as the former, or have moderate dependence on resources, and medium priority tasks are executed when resources allow; low priority tasks are background maintenance, data analysis or other non-immediate tasks, which can be processed using idle system resources without affecting the execution of high priority tasks.

[0094] Next, each computing node in the financial system is classified according to the real-time collected computing resource usage state data, and a node classification marker is obtained. The classification is mainly based on the load of the node, including CPU usage, memory occupancy, network bandwidth and other key indicators.

[0095] Idle nodes have sufficient resources and can immediately undertake new tasks, especially suitable for the rapid start of high priority tasks; lightly loaded nodes have a certain resource occupancy, but still have a certain remaining capacity that can be allocated, suitable for flexible execution of medium priority tasks, also called available nodes in the system; heavily loaded nodes are in a high resource occupancy state and are not suitable for immediately receiving new tasks, and need to wait for resource release or only execute low priority tasks to avoid performance degradation caused by overload, also called overloaded nodes in the system.

[0096] Finally, based on the task level marker and the node classification marker, resources are allocated to assign tasks of appropriate levels to corresponding state computing nodes, fully utilizing system resources while ensuring timely task response.

[0097] It should be noted that the priority matching principle provides that high priority tasks are allocated to idle or lightly loaded nodes first, medium priority tasks consider the resource utilization of lightly loaded nodes, and low priority tasks can be allocated to any non-heavy loaded node or wait for resource release of heavy loaded nodes before scheduling.

[0098] Through the above-mentioned hierarchical and classified resource allocation method, the embodiment of the present application brings significant technical effects in solving the technical problems: speeding up the response time of key tasks, improving resource utilization, and enhancing the flexibility of scheduling strategies. The task ranking and node classification resource allocation strategy of the embodiment of the present application not only solves the problems of uneven resource allocation and unintelligent scheduling decision, but also further improves the response speed, resource utilization efficiency and management simplicity of the system.

[0099] Optionally, in the financial database task scheduling method provided by the embodiment of the application, the step of classifying each database task in the task request queue based on the priority score to obtain a task level mark comprises: matching the priority score with a preset task classification threshold for each database task to obtain a first matching result, wherein the preset task classification threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; in the case that the first matching result indicates that the priority score is greater than or equal to the first threshold, marking the database task as a high-priority task; in the case that the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold, marking the database task as a medium-priority task; and in the case that the first matching result indicates that the priority score is less than the second threshold, marking the database task as a low-priority task.

[0100] Optionally, in the financial database task scheduling method provided by the embodiment of the application, the step of classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification mark comprises: matching the load evaluation value of the computing node indicated by the computing resource usage state with a preset node classification threshold for each computing node to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; in the case that the second matching result indicates that the load evaluation value is greater than or equal to the third threshold, marking the computing node as an overloaded node; in the case that the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold, marking the computing node as an available node; and in the case that the second matching result indicates that the load evaluation value is less than the fourth threshold, marking the computing node as an idle node.

[0101] A preferred embodiment is that the resource allocation strategy based on the task level mark and the node classification mark is a key step to ensure that different priority tasks are properly processed, and three allocation modes of resource monopoly, resource elasticity and resource limitation are used to provide optimal execution environments for high-priority tasks, medium-priority tasks and low-priority tasks respectively.

[0102] Optionally, in the task scheduling method of the financial database, the step of performing resource allocation based on the task level mark and the node classification mark to obtain a resource allocation result comprises: configuring a resource exclusive allocation strategy for the high-priority task, wherein the resource exclusive allocation strategy refers to preferentially allocating the high-priority task to an idle node, in the case that the idle node is fully occupied, allocating the remaining high-priority task to an available node, and prohibiting the high-priority task from being allocated to an overloaded node; configuring a resource elastic allocation strategy for the medium-priority task, wherein the resource elastic allocation strategy refers to, after the high-priority task is allocated, preferentially allocating the medium-priority task to the idle node in the case that there is an idle node that is not allocated, allocating the medium-priority task to the available node, and prohibiting the medium-priority task from being allocated to the overloaded node; configuring a resource limited allocation strategy for the low-priority task, wherein the resource limited allocation strategy refers to, after the high-priority task and the medium-priority task are both allocated, preferentially allocating the low-priority task to the idle node in the case that there is an idle node that is not allocated, allocating the low-priority task to the available node, and prohibiting the low-priority task from being allocated to the overloaded node; and performing resource allocation operation based on the resource exclusive allocation strategy, the resource elastic allocation strategy, and the resource limited allocation strategy to obtain the resource allocation result.

[0103] Specifically, the resource exclusive allocation strategy is adopted for the high-priority task, and the high-priority task is allocated to the idle node to provide sufficient resources to ensure its rapid execution. When the high-priority task is not allocated completely, the node with the most abundant resources is searched in the available node for allocation. Importantly, the high-priority task is never allocated to the overloaded node to avoid the influence of resource competition on the execution efficiency.

[0104] The resource elastic allocation strategy is adopted for the medium-priority task. After the resource allocation of the high-priority task is completed, the medium-priority task can be allocated only when there is an idle node that is not allocated in the system. The medium-priority task that fails to be allocated to the idle node is arranged to the available node, but the overloaded node is also avoided, thereby ensuring the execution of the high-priority task in priority and efficiently utilizing the system resources as much as possible to meet the execution requirement of the medium-priority task.

[0105] The resource limited allocation strategy is adopted for the low-priority task. The low-priority task can be allocated resources only when the high-priority task and the medium-priority task are both allocated and there is an idle node that is not used in the system. If the idle node is insufficient, the low-priority task is allocated to the available node that is not overloaded, thereby reasonably utilizing the system resources on the basis of ensuring the execution of the critical task in priority and avoiding the interference of the low-priority task on the high-priority task.

[0106] Step S204, performing resource scheduling on the task request queue based on the resource allocation result.

[0107] Preferably, step S204 not only involves reordering the task queue, but also includes intelligently allocating tasks to various computing nodes in the system and real-time adjustment of scheduling decisions.

[0108] Specifically, first, the task request queue is reordered according to the resource allocation result obtained in step S203. High-priority tasks are placed at the front of the queue to obtain resources and scheduling opportunities first, ensuring quick response to important and urgent tasks and meeting the real-time requirements of business.

[0109] Next, the reordered task queue is processed by an intelligent scheduling algorithm, which allocates tasks to the most suitable computing nodes according to the resource requirements of each task and the current state of the computing nodes, taking into account the load of the nodes, resource availability, and characteristics of the tasks, such as data affinity, preference for compute-intensive or I / O-intensive tasks, etc., to improve resource utilization and task execution efficiency.

[0110] In addition, the embodiments of the present application can also dynamically adjust the scheduling strategy to adapt to the real-time changes in resource usage and newly arrived task requests. For example, if a sudden increase in resource utilization of a node is detected, the task allocation is re-evaluated and some tasks are transferred to nodes with lower load to avoid resource bottlenecks, while high-priority tasks that have just arrived are quickly responded to and resources are re-allocated.

[0111] After performing resource scheduling, the execution status of the tasks and the resource usage of the computing nodes are continuously monitored for feedback optimization and adjustment of the scheduling strategy, ensuring the accuracy of the scheduling decisions and the adaptability of the system.

[0112] In addition, to handle emergency situations and resource conflicts, the embodiments of the present application implement pre-emptive and delayed scheduling mechanisms. Pre-emptive scheduling allows low-priority tasks to be interrupted when resources are scarce to ensure the execution of high-priority tasks, while delayed scheduling allows non-urgent tasks to be executed only when resources are available, reducing resource competition and maintaining system stability.

[0113] In summary, through task priority ordering based on resource allocation results, intelligent task allocation, dynamic strategy adjustment, resource monitoring feedback, scheduling mechanism optimization, and scheduling process auditing, the complex environment of high concurrency and multiple tasks can be effectively handled to ensure the smooth operation of critical business, improve data processing speed and overall system performance, and reduce operation and maintenance costs.

[0114] Through the above steps S201 to S204, the task request queue of the financial database can be received first, and the computing resource usage status of the financial system can be obtained. The task request queue contains N database tasks to be executed, where N is a positive integer. Then, the priority score of each database task in the task request queue is calculated. Then, computing resources are allocated to each database task based on the priority score and the computing resource usage status to obtain the resource allocation result. Finally, resource scheduling is performed on the task request queue based on the resource allocation result.

[0115] In this embodiment of the invention, a combination of intelligent priority evaluation and dynamic resource allocation is adopted. By integrating reinforcement learning algorithms and multi-dimensional task attribute analysis, the goal of accurately matching computing resources with task priority requirements is achieved. This results in a significant improvement in the task scheduling efficiency and resource utilization of financial databases. Specifically, after receiving the task request queue of the financial database, deep reinforcement learning technology is used to comprehensively analyze each database task to be executed, thereby accurately calculating the priority score of each task. Subsequently, based on real-time monitoring of computing resource usage and intelligent scheduling algorithms, a resource allocation strategy is dynamically generated to ensure that high-priority tasks can quickly obtain sufficient computing resources, while low-priority tasks will have the opportunity for reasonable scheduling and execution without affecting the overall system performance. This not only effectively avoids resource waste and computing bottlenecks but also significantly enhances the system's response speed and processing capacity. In particular, it can ensure the continuity and stability of core businesses during peak financial transaction periods, thereby solving the technical problem of low task resource scheduling efficiency in financial databases in related technologies.

[0116] The invention will now be described in conjunction with another alternative embodiment.

[0117] Example 2

[0118] This invention also provides a task scheduling device for a financial database. It should be noted that the task scheduling device for a financial database in this invention includes multiple implementation units, which can be used to execute the task scheduling method for a financial database provided in the first embodiment above. Each implementation unit corresponds to each implementation step in the first embodiment above.

[0119] Figure 3 This is a schematic diagram of an optional task scheduling device for a financial database according to an embodiment of the present invention, such as... Figure 3 As shown, the device may include: a receiving unit 31, a calculation unit 32, an allocation unit 33, and an execution unit 34.

[0120] The receiving unit 31 is configured to receive a task request queue of the financial database and obtain a computing resource usage state of the financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer.

[0121] The computing unit 32 is configured to calculate a priority score of each database task in the task request queue.

[0122] The allocating unit 33 is configured to allocate computing resources to each database task based on the priority score and the computing resource usage state to obtain a resource allocation result.

[0123] The executing unit 34 is configured to perform resource scheduling on the task request queue based on the resource allocation result.

[0124] The task scheduling device for the financial database can first receive the task request queue of the financial database through the receiving unit 31 and obtain the computing resource usage state of the financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer, then calculate the priority score of each database task in the task request queue through the computing unit 32, and then allocate computing resources to each database task based on the priority score and the computing resource usage state through the allocating unit 33 to obtain a resource allocation result, and finally perform resource scheduling on the task request queue based on the resource allocation result through the executing unit 34.

[0125] In the embodiment of the present application, the intelligent priority evaluation and dynamic resource allocation are combined, the reinforcement learning algorithm and multi-dimensional task attribute analysis are fused, the computing resource and the task priority requirement are accurately matched, the task scheduling efficiency and the resource utilization efficiency of the financial database are improved, and the technical effects are achieved. Specifically, after receiving the task request queue of the financial database, the deep reinforcement learning technology is used to analyze each database task to be executed, the priority score of each task is accurately calculated, the resource allocation strategy is dynamically generated based on the real-time monitoring of the computing resource usage state and the intelligent scheduling algorithm, the high-priority task can quickly obtain sufficient computing resources, the low-priority task can obtain reasonable scheduling and execution opportunities without affecting the overall efficiency of the system, the resource waste and the computing bottleneck are effectively avoided, the response speed and the processing capacity of the system are significantly improved, the continuity and the stability of the core business can be ensured during the peak period of financial transactions, and the technical problem of low task resource scheduling efficiency of the financial database in the related art is solved.

[0126] Further, the receiving unit comprises: a collecting module, configured to collect state parameters of M computing nodes in the financial system, wherein the state parameters comprise CPU usage, memory occupancy and network bandwidth occupancy, and M is a positive integer; a normalization processing module, configured to perform normalization processing on the collected state parameters to obtain resource index data in a standard format; a first weighting calculation module, configured to perform weighting calculation on resource index data of each computing node according to a first preset weight configuration to obtain a load evaluation value of each computing node; and a generating module, configured to generate a system-level resource usage state based on the load evaluation values of all computing nodes.

[0127] Further, the computing unit comprises: an obtaining module, configured to obtain task attribute parameters of each database task, wherein the task attribute parameters comprise an influence range parameter, a time limit requirement parameter and a resource demand parameter; and a second weighting calculation module, configured to perform weighting calculation on the influence range parameter, the time limit requirement parameter and the resource demand parameter according to a second preset weight configuration to obtain a priority score of each database task.

[0128] Further, the allocating unit comprises: a grading module, configured to grade each database task in the task request queue based on the priority score to obtain a task level mark; a classification module, configured to classify each computing node in the financial system based on the resource usage state to obtain a node classification mark; and an allocating module, configured to perform resource allocation based on the task level mark and the node classification mark to obtain a resource allocation result.

[0129] Further, the grading module comprises: a first matching submodule, configured to match the priority score with a preset task grading threshold for each database task to obtain a first matching result, wherein the preset task grading threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; a first marking submodule, configured to mark the database task as a high-priority task in a case where the first matching result indicates that the priority score is greater than or equal to the first threshold; a second marking submodule, configured to mark the database task as a medium-priority task in a case where the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold; and a third marking submodule, configured to mark the database task as a low-priority task in a case where the first matching result indicates that the priority score is less than the second threshold.

[0130] Further, the classification module comprises: a second distribution submodule, configured to, for each computing node, match the load evaluation value of the computing node indicated by the computing resource usage state with a preset node classification threshold to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; a fourth marking submodule, configured to, in a case where the second matching result indicates that the load evaluation value is greater than or equal to the third threshold, mark the computing node as an overload node; a fifth marking submodule, configured to, in a case where the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold, mark the computing node as an available node; and a sixth marking submodule, configured to, in a case where the second matching result indicates that the load evaluation value is less than the fourth threshold, mark the computing node as an idle node.

[0131] Further, the distribution module comprises: a first configuration submodule, configured to configure a resource exclusive distribution strategy for the high-priority tasks, wherein the resource exclusive distribution strategy refers to preferentially distributing the high-priority tasks to the idle nodes, in a case where the idle nodes are fully occupied, distributing the remaining high-priority tasks that are not distributed to the available nodes, and prohibiting the high-priority tasks from being distributed to the overload nodes; a second configuration submodule, configured to configure a resource elastic distribution strategy for the medium-priority tasks, wherein the resource elastic distribution strategy refers to, in a case where there are idle nodes that are not distributed after the high-priority tasks are distributed, preferentially distributing the medium-priority tasks to the idle nodes, distributing the medium-priority tasks that are not distributed to the available nodes, and prohibiting the medium-priority tasks from being distributed to the overload nodes; a third configuration submodule, configured to configure a resource limit distribution strategy for the low-priority tasks, wherein the resource limit distribution strategy refers to, in a case where there are idle nodes that are not distributed after the high-priority tasks and the medium-priority tasks are distributed, preferentially distributing the low-priority tasks to the idle nodes, distributing the low-priority tasks that are not distributed to the available nodes, and prohibiting the low-priority tasks from being distributed to the overload nodes; and an execution submodule, configured to perform a resource distribution operation based on the resource exclusive distribution strategy, the resource elastic distribution strategy, and the resource limit distribution strategy to obtain a resource distribution result.

[0132] It should be noted that the receiving unit 31, the computing unit 32, the distribution unit 33, and the execution unit 34 correspond to steps S201 to S204 in Embodiment One, and have the same instances and application scenarios as those of the corresponding steps, but are not limited to the content disclosed in Embodiment One. It should be noted that the modules or units can be hardware components or software components stored in a memory (for example, the memory 104) and processed by one or more processors (for example, the processors 102a, 102b, …, 102n), or can be a part of the apparatus and can run in the computer terminal 10 provided in Embodiment One.

[0133] The invention will now be described in conjunction with another alternative embodiment.

[0134] Example 3

[0135] The present invention can also provide an electronic device. Figure 4 This is a structural block diagram of an electronic device for executing a task scheduling method for a financial database according to an embodiment of the present invention, such as... Figure 4 As shown, the electronic device may include: one or more ( Figure 4 Only one of the following is shown: processor 402, memory 404, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0136] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the task scheduling method and apparatus for the financial database in this application embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned task scheduling method for the financial database. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0137] The processor can access information and applications stored in memory via a transmission device to perform the following steps: receiving a task request queue from a financial database and obtaining the computing resource usage status of the financial system, wherein the task request queue contains N database tasks to be executed, where N is a positive integer; calculating the priority score of each database task in the task request queue; allocating computing resources to each database task based on the priority score and the computing resource usage status to obtain the resource allocation result; and performing resource scheduling on the task request queue based on the resource allocation result.

[0138] The processor can also access information and applications stored in the memory via a transmission device to execute the following steps: collect status parameters of M computing nodes in the financial system, where the status parameters include: CPU utilization, memory usage, and network bandwidth usage, and M is a positive integer; normalize the collected status parameters to obtain resource indicator data in a standard format; perform weighted calculation on the resource indicator data of each computing node according to a first preset weight configuration to obtain the load assessment value of each computing node; and generate a system-level resource usage status based on the load assessment values ​​of all computing nodes.

[0139] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: for each database task, obtaining a task attribute parameter of the database task, wherein the task attribute parameter comprises an influence range parameter, a time limit requirement parameter and a resource demand parameter; performing weighted calculation on the influence range parameter, the time limit requirement parameter and the resource demand parameter according to a second preset weight configuration to obtain a priority score of each database task.

[0140] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: based on the priority score, classifying each database task in the task request queue to obtain a task level mark; based on the computing resource usage state, classifying each computing node in the financial system to obtain a node classification mark; and based on the task level mark and the node classification mark, performing resource allocation to obtain a resource allocation result.

[0141] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: for each database task, matching the priority score with a preset task classification threshold to obtain a first matching result, wherein the preset task classification threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; in a case where the first matching result indicates that the priority score is greater than or equal to the first threshold, marking the database task as a high-priority task; in a case where the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold, marking the database task as a medium-priority task; and in a case where the first matching result indicates that the priority score is less than the second threshold, marking the database task as a low-priority task.

[0142] The processor can further call information and application programs stored in the memory through the transmission device to perform the following steps: for each computing node, matching a load evaluation value of the computing node indicated by the computing resource usage state with a preset node classification threshold to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; in a case where the second matching result indicates that the load evaluation value is greater than or equal to the third threshold, marking the computing node as an overloaded node; in a case where the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold, marking the computing node as an available node; and in a case where the second matching result indicates that the load evaluation value is less than the fourth threshold, marking the computing node as an idle node.

[0143] The processor can also call information and application programs stored in the memory through the transmission device to perform the following steps: configuring a resource exclusive allocation strategy for high-priority tasks, wherein the resource exclusive allocation strategy refers to preferentially allocating high-priority tasks to idle nodes, allocating remaining high-priority tasks that are not allocated to available nodes in the case that the idle nodes are fully occupied, and prohibiting allocation of high-priority tasks to overloaded nodes; configuring a resource elastic allocation strategy for medium-priority tasks, wherein the resource elastic allocation strategy refers to preferentially allocating medium-priority tasks to idle nodes in the case that there are idle nodes that are not allocated after the high-priority tasks are allocated, allocating medium-priority tasks that are not allocated to available nodes, and prohibiting allocation of medium-priority tasks to overloaded nodes; configuring a resource limited allocation strategy for low-priority tasks, wherein the resource limited allocation strategy refers to preferentially allocating low-priority tasks to idle nodes in the case that there are idle nodes that are not allocated after the high-priority tasks and the medium-priority tasks are all allocated, allocating low-priority tasks that are not allocated to available nodes, and prohibiting allocation of low-priority tasks to overloaded nodes; and performing a resource allocation operation based on the resource exclusive allocation strategy, the resource elastic allocation strategy, and the resource limited allocation strategy to obtain a resource allocation result.

[0144] By adopting the embodiment of the present application, a task scheduling scheme of a financial database is provided. By combining intelligent priority evaluation and dynamic resource allocation, and by means of fusing a reinforcement learning algorithm and multi-dimensional task attribute analysis, the purpose of accurately matching computing resources and task priority requirements is achieved, thereby realizing the technical effects of significantly improving the efficiency of financial database task scheduling and the effectiveness of resource utilization. Specifically, after receiving a task request queue of the financial database, a deep reinforcement learning technology is used to comprehensively analyze each database task to be executed, thereby accurately calculating the priority score of each task. Then, based on the real-time monitoring of the computing resource usage state and the intelligent scheduling algorithm, a resource allocation strategy is dynamically generated to ensure that high-priority tasks can quickly obtain sufficient computing resources, while low-priority tasks can obtain reasonable scheduling and execution opportunities without affecting the overall effectiveness of the system. This not only effectively avoids resource waste and computing bottlenecks, but also significantly enhances the response speed and processing capacity of the system, especially during the peak period of financial transactions, which can guarantee the continuity and stability of core business, thereby solving the technical problem of low efficiency of task resource scheduling of the financial database in related technologies.

[0145] Those skilled in the art can understand that, Figure 4 The structure shown is only schematic, and the electronic device can also be a terminal device such as a smart phone, a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, and the like. Figure 4It does not cause limitation to the structure of the electronic device. For example, the electronic device can further include more or less components (such as a network interface, a display device, etc.) than those shown in FIG. 1, or have a different configuration from that shown in FIG. 1. Figure 4 Figure 4

[0146] Those skilled in the art can understand that all or part of the steps in the above-mentioned various methods of the embodiments can be completed by instructing the terminal device related hardware through a program, and the program can be stored in a computer readable storage medium, which can include a flash disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0147] The application will be described in detail below in combination with another alternative embodiment.

[0148] Embodiment Four

[0149] The embodiments of the application further provide a computer readable storage medium. Optionally, in the embodiments of the application, the computer readable storage medium can be used to save the program code executed by the task scheduling method of the financial database provided in the embodiment one.

[0150] Optionally, in the embodiments of the application, the storage medium can be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.

[0151] The embodiments of the application further provide a computer program product, when executed on a data processing device, is adapted to execute the steps of the task scheduling method of the financial database: receiving a task request queue of the financial database, and obtaining a computing resource usage state of the financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer; calculating a priority score of each database task in the task request queue; allocating computing resources for each database task based on the priority score and the computing resource usage state, to obtain a resource allocation result; and performing resource scheduling on the task request queue based on the resource allocation result.

[0152] The serial numbers of the embodiments of the application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0153] In the above embodiments of the application, the description of each embodiment has its own focus, and the parts not described in detail in an embodiment can be referred to the related description of other embodiments.

[0154] ​​In several embodiments provided in the present application, it should be understood that the disclosed technology can be implemented by other ways. Among them, the above-described device embodiments are only schematic, for example, the division of units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be indirect coupling or communication connection through some interfaces, units or modules, and can be electrical or other forms.

[0155] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e. they can be located in one place or distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0156] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0157] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0158] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A method for task scheduling of a financial database, characterized in that, The method comprises the following steps: receiving a task request queue of a financial database, and obtaining a computing resource usage state of a financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer; calculating a priority score of each database task in the task request queue; allocating computing resources to each database task based on the priority score and the computing resource usage state, to obtain a resource allocation result; performing resource scheduling on the task request queue based on the resource allocation result.

2. The task scheduling method of claim 1, wherein, The step of obtaining the computing resource usage state of the financial system comprises: collecting state parameters of M computing nodes in the financial system, wherein the state parameters comprise: CPU usage, memory occupancy, and network bandwidth occupancy, and M is a positive integer; performing normalization processing on the collected state parameters to obtain resource index data in a standard format; performing weighted calculation on the resource index data of each computing node according to a first preset weight configuration to obtain a load evaluation value of each computing node; generating a system-level resource usage state based on the load evaluation values of all computing nodes.

3. The task scheduling method of claim 1, wherein, The step of calculating the priority score of each database task in the task request queue comprises: for each database task, obtaining task attribute parameters of the database task, wherein the task attribute parameters comprise: an influence range parameter, a time limit requirement parameter, and a resource demand parameter; performing weighted calculation on the influence range parameter, the time limit requirement parameter, and the resource demand parameter according to a second preset weight configuration to obtain the priority score of each database task.

4. The task scheduling method of claim 1, wherein, The step of allocating computing resources to each database task based on the priority score and the computing resource usage state to obtain a resource allocation result comprises: grading each database task in the task request queue based on the priority score to obtain a task level mark; classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification mark; performing resource allocation based on the task level mark and the node classification mark to obtain the resource allocation result.

5. The task scheduling method of claim 4, wherein, The step of grading each database task in the task request queue based on the priority score to obtain a task level mark comprises: for each database task, matching the priority score with a preset task grading threshold to obtain a first matching result, wherein the preset task grading threshold comprises a first threshold and a second threshold, and the first threshold is greater than the second threshold; in a case where the first matching result indicates that the priority score is greater than or equal to the first threshold, marking the database task as a high-priority task; in a case where the first matching result indicates that the priority score is less than the first threshold and greater than or equal to the second threshold, marking the database task as a medium-priority task; in a case where the first matching result indicates that the priority score is less than the second threshold, marking the database task as a low-priority task.

6. The task scheduling method of claim 4, wherein, The step of classifying each computing node in the financial system based on the computing resource usage state to obtain a node classification label comprises: For each computing node, a load evaluation value of the computing node indicated by the computing resource usage state is matched with a preset node classification threshold to obtain a second matching result, wherein the preset node classification threshold comprises a third threshold and a fourth threshold, and the third threshold is greater than the fourth threshold; In a case where the second matching result indicates that the load evaluation value is greater than or equal to the third threshold, the computing node is marked as an overload node; In a case where the second matching result indicates that the load evaluation value is less than the third threshold and greater than or equal to the fourth threshold, the computing node is marked as an available node; In a case where the second matching result indicates that the load evaluation value is less than the fourth threshold, the computing node is marked as an idle node.

7. The task scheduling method of claim 4, wherein, The step of performing resource allocation based on the task level label and the node classification label to obtain a resource allocation result comprises: A resource exclusive allocation strategy is configured for a high-priority task, wherein the resource exclusive allocation strategy refers to preferentially allocating the high-priority task to an idle node, in a case where the idle node is fully occupied, allocating the remaining high-priority tasks that are not allocated to an available node, and prohibiting the high-priority task from being allocated to an overload node; A resource elastic allocation strategy is configured for a medium-priority task, wherein the resource elastic allocation strategy refers to, after the high-priority task allocation is completed, in a case where there is an idle node that is not allocated, preferentially allocating the medium-priority task to the idle node, allocating the medium-priority tasks that are not allocated to the available node, and prohibiting the medium-priority task from being allocated to the overload node; A resource limit allocation strategy is configured for a low-priority task, wherein the resource limit allocation strategy refers to, after the high-priority task and the medium-priority task are all allocated, in a case where there is an idle node that is not allocated, preferentially allocating the low-priority task to the idle node, allocating the low-priority tasks that are not allocated to the available node, and prohibiting the low-priority task from being allocated to the overload node; A resource allocation operation is performed based on the resource exclusive allocation strategy, the resource elastic allocation strategy and the resource limit allocation strategy to obtain the resource allocation result.

8. An apparatus for task scheduling of a financial database, characterized by comprising: Comprise: A receiving unit is configured to receive a task request queue of a financial database and obtain a computing resource usage state of a financial system, wherein the task request queue contains N database tasks to be executed, and N is a positive integer; A computing unit is configured to calculate a priority score of each database task in the task request queue; An allocation unit is configured to allocate computing resources for each database task based on the priority score and the computing resource usage state to obtain a resource allocation result; An execution unit is configured to perform resource scheduling on the task request queue based on the resource allocation result.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to perform the task scheduling method of the financial database in any one of claims 1 to 7 when the computer program is running.

10. An electronic device, comprising: The device comprises one or more processors and a memory, and the memory is used to store one or more programs, wherein the one or more programs make the one or more processors implement the task scheduling method of the financial database in any one of claims 1 to 7 when the one or more programs are executed by the one or more processors.

11. A computer program product, characterised in that, The device comprises computer instructions, wherein the computer instructions implement the steps of the task scheduling method of the financial database in any one of claims 1 to 7 when the computer instructions are executed by a processor.

Images