Resource scheduling distribution management method and system, medium and processor

Abstract

The invention relates to the technical field of resource scheduling, and discloses a resource scheduling distribution management method and system, a medium and a processor, and the method comprises the steps: obtaining heterogeneous resource information of each node under a system computing power network; acquiring quotas of various heterogeneous resources applied by the configuration file of the Pod to be scheduled in the system; according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied by the to-be-scheduled Pod, calculating a node load score and a node resource balance degree score of the to-be-scheduled Pod scheduled to each node; obtaining a scheduling scoring result of each node according to the node load score and the node resource balance degree score; and according to the scheduling score result, selecting the node with the highest scheduling score in all the nodes to schedule the current Pod to be scheduled. According to the method, two factors of node load and node resource balance degree can be comprehensively considered, various heterogeneous resources under the computing power network are managed and scheduled, and the load of each node and dynamic allocation of the resources are balanced.

Description

Technical Field

[0001] The present invention relates to the technical field of resource scheduling, and in particular to a resource scheduling allocation management method, system, medium and processor. Background Art

[0002] The development of the digital economy will drive the generation of massive amounts of data, and data processing requires powerful computing power coordinated by the cloud, edge, and end, as well as extensive network connections. Diversified computing power and computing-network integration will become important trends.

[0003] In practical applications, diverse computing power and integrated computing power networks usually need to support a variety of different algorithm requirements, including machine learning, deep learning and other algorithms. In order to meet these algorithm requirements, diverse computing power and integrated computing power networks need to support different types of heterogeneous resources, such as CPU, GPU, FPGA, etc., as well as high-speed network and storage devices to provide more efficient computing and communication performance. At the same time, in terms of resource management, the construction of heterogeneous computing power resources also needs to consider efficient resource management strategies. This makes how to balance the scheduling of multiple heterogeneous computing power resources in the computing power network an urgent problem to be solved.

[0004] In view of this, it is necessary to provide a resource scheduling and allocation management method, system, medium and processor. Summary of the invention

[0005] In view of the problem of how to balance and schedule various heterogeneous computing resources in computing power networks in the prior art, the present invention provides a resource scheduling allocation management method, system, medium and processor, which can comprehensively consider the factors of node load and node resource balance, and provide a solution for the balanced scheduling of multi-dimensional heterogeneous resources in computing power networks. The specific technical solution is as follows:

[0006] A resource scheduling and allocation management method, comprising:

[0007] Obtain heterogeneous resource information of each node in the system computing network;

[0008] Get the quotas of various heterogeneous resources applied for by the configuration files of the Pods to be scheduled in the system;

[0009] Based on the heterogeneous resource information of each node and the quotas of various heterogeneous resources applied for by the Pod to be scheduled, the node load score and node resource balance score of the Pod to be scheduled to each node are calculated;

[0010] Obtaining a scheduling score result for each node according to the node load score and the node resource balance score;

[0011] According to the scheduling score result, the node with the highest scheduling score among all nodes is selected to schedule the current Pod to be scheduled.

[0012] Preferably, the obtaining of resource information of each node in the system computing network includes:

[0013] Obtain the resource optimization plug-in deployed on each node in the system computing network;

[0014] Registering corresponding heterogeneous resources on each node according to the resource optimization plug-in;

[0015] By monitoring the heterogeneous resources registered by the resource optimization plug-in, the heterogeneous resource information of each node in the system computing power network is obtained.

[0016] Preferably, a resource scheduling and allocation management method further includes:

[0017] The heterogeneous resource information is analyzed and processed to obtain the total capacity of heterogeneous resources of each node in the system computing network, the number of heterogeneous resource types and the initial usage of each type of heterogeneous resources.

[0018] Preferably, the calculating the node load score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of each type of heterogeneous resources applied for by the Pod to be scheduled includes:

[0019] According to the quotas of various heterogeneous resources applied by the Pod to be scheduled and the initial usage of various heterogeneous resources, calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node;

[0020] Calculate the remaining resources of each node based on the total capacity of heterogeneous resources and the current usage of each type of heterogeneous resources;

[0021] The node load score of each node is calculated based on the number of heterogeneous resource types, the total capacity of heterogeneous resources, the remaining resources and the preset weight coefficients of various types of heterogeneous resources.

[0022] Preferably, the calculating the node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of each type of heterogeneous resources applied for by the Pod to be scheduled includes:

[0023] According to the quotas of various heterogeneous resources applied by the Pod to be scheduled and the initial usage of various heterogeneous resources, calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node;

[0024] Calculate the resource utilization of each node based on the total capacity of the heterogeneous resources of each node and the current usage of each type of heterogeneous resources;

[0025] The resource balance score of each node is obtained based on the resource utilization of each node and the standard deviation of the resource utilization of various heterogeneous resources.

[0026] Preferably, the step of obtaining the quotas of various heterogeneous resources applied for by the configuration file of the Pod to be scheduled in the system includes:

[0027] Get the list information of all Pods in the system, including name and status information;

[0028] By traversing the list information of all Pods, filter out the Pods to be scheduled that contain the names to be scheduled;

[0029] By searching for quota information in the configuration file of the filtered Pod to be scheduled, the quota information of various heterogeneous resources applied for by the Pod to be scheduled is obtained.

[0030] Preferably, after selecting the node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result, the method further includes:

[0031] Bind the Pod to be scheduled to the selected node and generate scheduling information;

[0032] Update the status of the Pod according to the scheduling information, and schedule the Pod to be scheduled to the bound node to run the Pod.

[0033] A resource scheduling and allocation management system, applied to the aforementioned resource scheduling and allocation management method, comprises:

[0034] A resource information acquisition unit, used to acquire heterogeneous resource information of each node in the system computing network;

[0035] The Pod information acquisition unit is used to obtain the quotas of various heterogeneous resources applied for by the configuration files of the Pods to be scheduled in the system;

[0036] The balance score calculation unit is used to calculate the node load score and node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled;

[0037] A scheduling score calculation unit, used to obtain a scheduling score result of each node according to the node load score and the node resource balance score;

[0038] The resource scheduling selection unit is used to select a node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result.

[0039] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the aforementioned resource scheduling allocation management method.

[0040] A processor is used to run a program, wherein the program executes the above-mentioned resource scheduling allocation management method when running.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] A resource scheduling and allocation management method of the present invention obtains the heterogeneous resource information of each node under the system computing power network; obtains the quota of various heterogeneous resources applied for by the configuration file of the Pod to be scheduled in the system; calculates the node load score and node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled; obtains the scheduling score result of each node according to the node load score and the node resource balance score; according to the scheduling score result, selects the node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled. The present invention can comprehensively consider the two factors of node load and node resource balance, and uniformly manage and schedule multi-dimensional heterogeneous resources under the computing power network, realizes balancing the load of each node and dynamically allocating resources, and improves the utilization rate of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0044] Figure 1 The present invention is a flow chart of a resource scheduling and allocation management method.

[0045] Figure 2 The present invention is a flowchart of an embodiment of a resource scheduling and allocation management method of the present invention.

[0046] Figure 3 The present invention is a flow chart of another embodiment of a resource scheduling and allocation management method of the present invention.

[0047] Figure 4 The present invention is a flow chart of another embodiment of a resource scheduling and allocation management method.

[0048] Figure 5 This is a schematic diagram of a resource scheduling and allocation management system of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0051] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0052] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0053] The scheduling of heterogeneous resources in the existing system computing network is mainly implemented by the Kubernetes resource orchestration management platform. However, the current native scheduling mechanism of open source Kubernetes generally only considers two resources, CPU and memory, in terms of resource dimensions, and does not take into account ubiquitous heterogeneous computing resources such as GPU and FGPA in the computing network. Therefore, the native scheduling strategy cannot meet the multi-dimensional heterogeneous resource scheduling requirements in the computing network. The present application discloses a resource scheduling and allocation management method that is optimized on the Kubernetes resource orchestration management platform, registers corresponding heterogeneous resources on each node in the system, realizes the abstraction of heterogeneous resources and realizes unified management, and realizes balanced scheduling by comprehensively considering the cluster load and the balance of heterogeneous resources.

[0054] See the following examples Figures 1 to 5 .

[0055] The present application provides a method for scheduling and allocating resources, including:

[0056] Step S01, obtaining heterogeneous resource information of each node in the system computing network;

[0057] Heterogeneous resources under the computing network include general resources such as CPU, storage, network bandwidth, and AI accelerators such as GPU and FPG. Different heterogeneous resources have different computing capabilities and characteristics and are suitable for different computing tasks. For example, CPU is suitable for general computing tasks, while GPU is good at parallel computing.

[0058] Specifically, the acquisition of resource information of each node in the system computing network includes:

[0059] Step S11, obtaining the resource optimization plug-in deployed on each node in the system computing network;

[0060] First, deploy the implemented custom resource optimization plug-in (such as deviceplugin or daemonset) to the Kubernetes cluster.

[0061] Step S12: registering corresponding heterogeneous resources on each node according to the resource optimization plug-in;

[0062] Each node in the system needs to deploy a resource optimization plug-in to register heterogeneous resources on this node. The registered information mainly includes APIVersion, ResourceName, and unixsocket of deviceplugin.

[0063] Step S13: Obtain heterogeneous resource information of each node in the system computing network by monitoring the heterogeneous resources registered by the resource optimization plug-in.

[0064] After deviceplugin successfully registers with kubelet, kubelet monitors the current plug-in registration resources through the ListAndWatch interface.

[0065] Deployed to the scheduler in the form of a plug-in, the corresponding optimization technology is used to abstract heterogeneous resources into abstract resources that can be uniformly scheduled and managed and registered into the cluster.

[0066] The heterogeneous resource information is analyzed and processed to obtain the total capacity of heterogeneous resources of each node in the system computing network, the number of heterogeneous resource types and the initial usage of each type of heterogeneous resources.

[0067] Step S02: Obtain the quotas of various heterogeneous resources applied for by the configuration file of the Pod to be scheduled in the system;

[0068] Before getting the Pod configuration file, you need to install and configure kubectl. kubectl is a command-line tool for Kubernetes that is used to interact with the Kubernetes cluster. Connect to the Kubernetes cluster by connecting to it, where "context" is the name of the context set when configuring kubectl.

[0069] Specifically, the quotas of various heterogeneous resources applied for by the configuration files of the Pod to be scheduled in the system include:

[0070] Step S21, obtain the list information of all Pods in the system, including name and status information;

[0071] Use the kubectl get pods command to obtain a list of all Pods, including name, status, restart count, etc. The name information can be used to identify which Pods are currently in the waiting-to-be-scheduled state.

[0072] Step S22: traverse the list information of all Pods to filter out the Pods to be scheduled that contain the names to be scheduled;

[0073] For example, if the Pod is currently in the pending state and its name is Pending, filter out the Pods in the Pending state from the obtained Pod list. These Pods are the Pods to be scheduled. <pod-name>,Use kubectl getpod <pod-name>The -oyaml command gets the detailed configuration file for a specific Pod.

[0074] Step S23: by searching the quota information in the configuration file of the filtered Pod to be scheduled, the quota information of various types of heterogeneous resources applied for by the Pod to be scheduled is obtained.

[0075] In the Pod configuration file, find the resources field under the spec.containers section. The field contains quota information for various heterogeneous resources requested by the Pod. According to the naming conventions of heterogeneous resources, extract the specific quotas of the heterogeneous resources requested by the Pod, including the requested amount and the limited amount.

[0076] Step S03: Calculate the node load score and node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of each type of heterogeneous resources applied for by the Pod to be scheduled;

[0077] Balanced scheduling is achieved by comprehensively considering cluster load and heterogeneous resource balancing, and the node load score and node resource balance score of each node are calculated respectively.

[0078] Assume that node N i The total capacity of heterogeneous resources is Where i = {1, 2, 3…x} represents the node number, j = {1, 2, 3…y} represents the type of heterogeneous resources; the initial usage of resources on node i is represented by Y, represents the initial usage of heterogeneous resource j on node i.

[0079] Specifically, the node load score of the Pod to be scheduled to each node is calculated based on the heterogeneous resource information of each node and the quota of each type of heterogeneous resources applied for by the Pod to be scheduled, including:

[0080] (1) Based on the quota R of various heterogeneous resources requested by the Pod to be scheduled j (j={1,2,3…y}) and the initial usage of various heterogeneous resources Calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node

[0081]

[0082] (2) Based on the total capacity of heterogeneous resources and the current usage of various heterogeneous resources Calculate the remaining resources of each node

[0083]

[0084] (3) Based on the number of heterogeneous resource types y and the total capacity of heterogeneous resources at each node Remaining resources The node load score of each node is calculated by using the preset weight coefficient λ of various heterogeneous resources.

[0085]

[0086] By calculating the percentage of the remaining resources of each heterogeneous resource to the total amount of heterogeneous resources, the remaining resources of each heterogeneous resource on the node are comprehensively considered, and the Pod is scheduled to the node with higher remaining resources. At the same time, considering the heterogeneity of resources, a weight coefficient is set for each heterogeneous resource to indicate the importance of various heterogeneous resources. The weight system can be set according to actual usage experience. The larger the weight coefficient, the priority is given to the balance of this heterogeneous resource. For example, if you want CPU resources to be more evenly distributed in the entire cluster, set the weight coefficient a little larger, so that the higher the remaining ratio of the node CPU, the higher the score.

[0087] Specifically, the quota R of various heterogeneous resources applied for by the Pod to be scheduled is based on the heterogeneous resource information of each node and j , the node resource balance score of the Pod to be scheduled to each node is calculated including:

[0088] (a) Based on the quota R of various heterogeneous resources requested by the Pod to be scheduled j and the initial usage of various heterogeneous resources Calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node

[0089]

[0090] (b) Based on the total capacity of heterogeneous resources at each node and the current usage of various heterogeneous resources Calculate the resource utilization of each node;

[0091]

[0092] (c) Based on the resource utilization U of each node i And the standard deviation σ corresponding to the resource utilization of various heterogeneous resources, the resource balance score of each node is obtained

[0093]

[0094] The lower the resource balance score, the more balanced the node resources are. This score tends to schedule Pods to nodes with more even resource usage, thereby improving resource utilization.

[0095] Step S04: Obtain the scheduling score result of each node according to the node load score and the node resource balance score

[0096]

[0097] Step S05: According to the scheduling score result, the node with the highest scheduling score among all nodes is selected to schedule the current Pod to be scheduled.

[0098] Specifically, after selecting the node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result, the method further includes:

[0099] Bind the Pod to be scheduled to the selected node and generate scheduling information;

[0100] Update the status of the Pod according to the scheduling information, and schedule the Pod to be scheduled to the bound node to run the Pod.

[0101] By scheduling Pods to the most suitable nodes, you can make full use of the resources in the cluster, avoid waste and idleness of resources, reduce the operating costs of the cluster, and improve resource utilization.

[0102] The resource scheduling and allocation management method of the present invention comprehensively considers two factors: node load and node resource balance, and uniformly manages and schedules multi-dimensional heterogeneous resources in a computing power network, thereby balancing the load of each node and dynamically allocating resources, thereby improving computing resource utilization.

[0103] The embodiment of the present application further provides a resource scheduling and allocation management system, which is applied to the aforementioned resource scheduling and allocation management method, including:

[0104] A resource information acquisition unit, used to acquire heterogeneous resource information of each node in the system computing network;

[0105] The Pod information acquisition unit is used to obtain the quotas of various heterogeneous resources applied for by the configuration files of the Pods to be scheduled in the system;

[0106] The balance score calculation unit is used to calculate the node load score and node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled;

[0107] A scheduling score calculation unit, used to obtain a scheduling score result of each node according to the node load score and the node resource balance score;

[0108] The resource scheduling selection unit is used to select a node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result.

[0109] Specifically, a resource scheduling and allocation management system also includes:

[0110] The resource information analysis and processing unit is used to analyze and process the heterogeneous resource information to obtain the total capacity of heterogeneous resources, the number of heterogeneous resource types and the initial usage of each type of heterogeneous resources of each node in the system computing network.

[0111] The functional explanation of each unit of a resource scheduling and allocation management system of the present application is the same as that of an embodiment of a resource scheduling and allocation management method, and the beneficial effects are also the same, so they will not be repeated here.

[0112] An embodiment of the present application provides a computer-readable storage medium, which includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the aforementioned resource scheduling and allocation management method.

[0113] The technical effect of this embodiment is the same as the technical effect of a resource scheduling allocation management method in an embodiment, and will not be repeated here.

[0114] The present invention can be used in many general or special computer system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like.

[0115] An embodiment of the present application provides a processor, which is used to run a program, wherein the program executes the aforementioned resource scheduling allocation management method when running.

[0116] The technical effect of this embodiment is the same as the technical effect of the resource scheduling allocation management method in Embodiment 1, and will not be repeated here.

[0117] The processor in this embodiment may be a central processing unit (CPU), a controller, a microcontroller, or other data processing chips.

[0118] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0119] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0120] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0121] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A resource scheduling and allocation management method, characterized in that: include: Obtain heterogeneous resource information of each node in the system computing network; Get the quotas of various heterogeneous resources applied for by the configuration files of the Pods to be scheduled in the system; Based on the heterogeneous resource information of each node and the quotas of various heterogeneous resources applied for by the Pod to be scheduled, the node load score and node resource balance score of the Pod to be scheduled to each node are calculated; Obtaining a scheduling score result for each node according to the node load score and the node resource balance score; According to the scheduling score result, the node with the highest scheduling score among all nodes is selected to schedule the current Pod to be scheduled.

2. A resource scheduling and allocation management method according to claim 1, characterized in that: The acquisition of resource information of each node in the system computing network includes: Obtain the resource optimization plug-in deployed on each node in the system computing network; Registering corresponding heterogeneous resources on each node according to the resource optimization plug-in; By monitoring the heterogeneous resources registered by the resource optimization plug-in, the heterogeneous resource information of each node in the system computing power network is obtained.

3. A resource scheduling and allocation management method according to claim 2, characterized in that: Also includes: The heterogeneous resource information is analyzed and processed to obtain the total capacity of heterogeneous resources of each node in the system computing network, the number of heterogeneous resource types and the initial usage of each type of heterogeneous resources.

4. A resource scheduling and allocation management method according to claim 3, characterized in that: The calculation of the node load score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled includes: According to the quotas of various heterogeneous resources applied by the Pod to be scheduled and the initial usage of various heterogeneous resources, calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node; Calculate the remaining resources of each node based on the total capacity of heterogeneous resources and the current usage of each type of heterogeneous resources; The node load score of each node is calculated based on the number of heterogeneous resource types, the total capacity of heterogeneous resources, the remaining resources and the preset weight coefficients of various types of heterogeneous resources.

5. The resource scheduling and allocation management method according to claim 3, characterized in that: The calculation of the node resource balance score of the Pod to be scheduled to each node based on the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled includes: According to the quotas of various heterogeneous resources applied by the Pod to be scheduled and the initial usage of various heterogeneous resources, calculate the current usage of various heterogeneous resources after the Pod to be scheduled is scheduled to each node; The resource utilization of each node is calculated based on the total capacity of the heterogeneous resources of each node and the current usage of each type of heterogeneous resources. The resource balance score of each node is obtained based on the resource utilization of each node and the standard deviation corresponding to the resource utilization of each type of heterogeneous resources.

6. The resource scheduling and allocation management method according to claim 3, characterized in that: The quotas of various heterogeneous resources applied for by the configuration files of the Pod to be scheduled in the system include: Get the list information of all Pods in the system, including name and status information; By traversing the list information of all Pods, filter out the Pods to be scheduled that contain the names to be scheduled; By searching for quota information in the configuration file of the filtered Pod to be scheduled, the quota information of various heterogeneous resources applied for by the Pod to be scheduled is obtained.

7. The method for resource scheduling and allocation management according to claim 3, characterized in that: After selecting the node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result, the method further includes: Bind the Pod to be scheduled to the selected node and generate scheduling information; Update the status of the Pod according to the scheduling information, and schedule the Pod to be scheduled to the bound node to run the Pod.

8. A resource scheduling and allocation management system, characterized in that: A resource scheduling and allocation management method applied to any one of claims 1 to 7, comprising: A resource information acquisition unit, used to acquire heterogeneous resource information of each node in the system computing network; The Pod information acquisition unit is used to obtain the quotas of various heterogeneous resources applied for by the configuration files of the Pods to be scheduled in the system; The balance score calculation unit is used to calculate the node load score and node resource balance score of the Pod to be scheduled to each node according to the heterogeneous resource information of each node and the quota of various heterogeneous resources applied for by the Pod to be scheduled; A scheduling score calculation unit, used to obtain a scheduling score result of each node according to the node load score and the node resource balance score; The resource scheduling selection unit is used to select a node with the highest scheduling score among all nodes to schedule the current Pod to be scheduled according to the scheduling score result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the resource scheduling and allocation management method according to any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the resource scheduling allocation management method described in any one of claims 1 to 7.

Images