Basic environment arrangement method and device based on cloud platform, equipment and medium

Abstract

The invention provides a basic environment arrangement method and device based on a cloud platform, equipment and a medium, and relates to the field of data processing, in particular to the technical field of cloud platforms. The specific implementation scheme is as follows: obtaining function demand information and a target network segment of a user for issuing an execution task to a cloud platform; determining corresponding platform resource information from the cloud platform according to the function demand information; and generating a basic environment arrangement result according to the platform resource information and a plurality of subnets obtained by dividing the target network segment, and providing the basic environment arrangement result to the user for task execution. According to the scheme provided by the invention, the user can execute the task meeting the user requirement without paying attention to the bottom layer resource composition of the cloud platform, so that the task attention of the user and the task execution experience on the cloud platform are improved; available resources on the cloud platform are effectively integrated, and the resource utilization rate of the cloud platform is increased.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing, in particular to the field of cloud platform technology, and specifically to a basic environment orchestration method, device, equipment and medium based on a cloud platform. Background Art

[0002] With the continuous development of AI technology, when users need to perform complex large-model training and inference tasks, they need to use the powerful underlying computing power, high-speed network and high-performance storage provided by different platforms to build a large-model pre-training environment. Faced with the diverse and complex resource and functional requirements of different users, how to improve the efficiency of building the basic environment for users has become an urgent problem to be solved. Summary of the invention

[0003] The present disclosure provides a method, apparatus, device and medium for orchestrating a basic environment based on a cloud platform.

[0004] According to one aspect of the present disclosure, a basic environment orchestration method based on a cloud platform is provided, comprising:

[0005] Obtain the functional requirement information and target network segment of the user for the cloud platform to issue execution tasks;

[0006] According to the functional requirement information, determining corresponding platform resource information from the cloud platform;

[0007] A basic environment orchestration result is generated according to the platform resource information and a plurality of subnets obtained by dividing the target network segment, and the basic environment orchestration result is provided to a user for task execution.

[0008] According to another aspect of the present disclosure, a basic environment orchestration device based on a cloud platform is provided, comprising:

[0009] Functional requirement acquisition module, used to obtain the functional requirement information and target network segment of the user for the cloud platform to issue execution tasks;

[0010] A platform resource determination module, used to determine corresponding platform resource information from the cloud platform according to the functional requirement information;

[0011] The environment orchestration module is used to generate a basic environment orchestration result according to the platform resource information and a plurality of subnets obtained by dividing the target network segment, and provide the basic environment orchestration result to the user for task execution.

[0012] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the cloud platform-based basic environment orchestration method described in any embodiment of the present disclosure.

[0016] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the cloud platform-based basic environment orchestration method according to any embodiment of the present disclosure.

[0017] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the cloud platform-based basic environment orchestration method according to any embodiment of the present disclosure.

[0018] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0020] Figure 1 is a schematic diagram of a basic environment orchestration method based on a cloud platform according to an embodiment of the present disclosure;

[0021] Figure 2 is a schematic diagram of another basic environment orchestration method based on a cloud platform according to an embodiment of the present disclosure;

[0022] Figure 3 It is an architecture diagram based on the basic environment orchestration method of the cloud platform;

[0023] Figure 4 It is a schematic diagram of the interconnection of cluster heterogeneous computing resource pools;

[0024] Figure 5 is a schematic diagram of another basic environment orchestration method based on a cloud platform according to an embodiment of the present disclosure;

[0025] Figure 6 This is a diagram of the target network segment division;

[0026] Figure 7 is a structural diagram of a basic environment orchestration device based on a cloud platform according to an embodiment of the present disclosure;

[0027] Figure 8It is a block diagram of an electronic device used to implement the cloud platform-based basic environment orchestration method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0028] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0029] Figure 1 This is a schematic diagram of a basic environment arrangement method based on a cloud platform according to an embodiment of the present disclosure. This embodiment is applicable to the situation where a basic environment corresponding to an execution task submitted by a user on a cloud platform is arranged. This method can be executed by a basic environment arrangement device based on a cloud platform. The device can be implemented by software and / or hardware and integrated in an electronic device. The electronic device involved in this embodiment can be a server with computing and communication capabilities, etc. Specifically, refer to Figure 1 , the method specifically includes the following:

[0030] S110, obtaining the functional requirement information and target network segment of the user for the cloud platform to issue the execution task.

[0031] Since users need to use the powerful underlying computing power, high-speed network and high-performance storage capabilities of the cloud platform to perform tasks, such as training or reasoning tasks of large models, the cloud platform integrates the central cloud server and the edge server provided by the third-party platform. Different types of computing cards are integrated in the central cloud server and the edge server to meet the different execution tasks of users.

[0032] Specifically, the user determines the functional requirement information for executing the task, such as the model framework that the user needs to train, and the required computing resource list and network requirements, etc., where the computing resource list includes the machine type and quantity of computing resources that the user needs to use. Exemplarily, the cloud platform provides a variety of candidate functional requirement information, and the candidate functional requirement information is determined based on the attribute information of the cloud platform itself. The user selects the corresponding target functional requirement information from the candidate functional requirement information based on the execution task and his own needs.

[0033] The target network segment provided by the user is the network information provided to the cloud platform for access in the user's private network. The cloud platform can access the resource data in the user's private network through the IP in the target network segment to perform tasks based on the resource data obtained from the access.

[0034] The cloud platform can provide the basic environment for large-model pre-training required by users. Users can take advantage of the powerful underlying computing power, high-speed network, and high-performance storage provided by the cloud platform to build a large-model pre-training environment in one stop, supporting a variety of mainstream training frameworks. At the same time, they can use the AI ​​acceleration capabilities provided by the cloud platform to accelerate the large-model pre-training process.

[0035] S120. Determine corresponding platform resource information from the cloud platform according to the functional requirement information.

[0036] Functional requirement information is the requirement information determined by the user from the perspective of the functional realization of the task and the actual cost control. According to the contents included in the functional requirement information, the matching platform resource information is determined from the cloud platform, so that the determined platform resource information can not only realize the functional realization of the task required by the user, but also meet the actual cost control conditions required by the user. The platform resource information includes various infrastructures required for the execution of user tasks, such as computing resources, storage and network.

[0037] Exemplarily, matching candidate machines are determined from the cloud platform based on the machine type and quantity in the functional requirement information, and then the corresponding target machine is determined from the candidate machines in combination with the user's cost control conditions as the underlying computing resources for executing user tasks; and based on the network and storage requirements in the functional requirements, the corresponding target network resources and target storage resources are determined from the cloud platform, and the platform resource information is determined based on the underlying computing resources, target network resources, and target storage resources.

[0038] In another optional implementation of this embodiment, S120 includes:

[0039] Decompose the execution task according to the functional requirement information to obtain multiple subtasks;

[0040] Determine the corresponding target platform resources from the candidate platform resources of the cloud platform for each subtask.

[0041] Since the cloud platform includes powerful underlying computing resources, in order to improve the efficiency of task execution, the task can be decomposed to obtain multiple subtasks, and the multiple subtasks can be assigned to multiple devices for training to improve the utilization of the underlying computing resources in the cloud platform and the execution efficiency of user tasks.

[0042] Specifically, the execution task is decomposed according to the user's functional requirement information so that the decomposed subtasks can be processed in parallel to the greatest extent possible to improve the efficiency of task execution. Exemplarily, the execution order is divided according to the user's functional requirement information, and the corresponding subtasks are determined according to the divided sub-functions.

[0043] According to the specific execution content of each subtask, the corresponding target platform resources are determined from the candidate platform resources of the cloud platform. Exemplarily, according to the specific execution content of each subtask, multiple matching platform resources are determined from the currently idle candidate platform resources in the cloud platform, and then combined with the target functional requirements associated with the subtask in the user's functional requirement information, the target platform resources corresponding to the subtask are determined from multiple platform resources. Among them, according to the specific execution content of each subtask, multiple matching platform resources are determined from the currently idle candidate platform resources in the cloud platform, which can be considered from the perspective of task execution efficiency.

[0044] This embodiment decomposes the execution task and determines the corresponding target platform resources according to the decomposed subtasks, refines the granularity of the target platform resource determination basis, improves the match between the target platform resources and the user's functional requirements, and thereby improves the accuracy and efficiency of the target platform resource determination.

[0045] S130: Generate a basic environment orchestration result according to the platform resource information and multiple subnets obtained by dividing the target network segment, and provide the basic environment orchestration result to the user for task execution.

[0046] When the cloud platform executes tasks, different tasks will simultaneously access the target network segment provided by the user to obtain data resources stored in the user's private network. Therefore, in order to avoid access conflicts, the target network segment is divided into multiple subnets. Different tasks can access the user's private network through different subnets.

[0047] The platform resource information determined in the cloud platform and the relationship between multiple subnets are orchestrated and integrated to generate the basic environment orchestration results required for user task execution. The cloud platform provides the orchestrated basic environment to the user, allowing the user to directly execute tasks through the orchestrated basic environment.

[0048] Users execute tasks through the basic environment orchestration results, so that users only need to pay attention to the execution progress of the task itself, without having to manage the underlying basic platform resources, such as the specific machine types and quantities used; and the basic environment orchestration results can meet users' diverse and complex resource and functional requirements, while being able to integrate available resources within the cloud platform, thereby improving the resource utilization efficiency within the cloud platform.

[0049] The solution of this embodiment determines the corresponding platform resource information through the functional requirement information of the task to be executed, and divides the target network segment provided by the user, and then generates the basic environment orchestration result of the task execution according to the platform resource information and the subnet obtained by the division, so that the user can execute the task that meets the user's needs without paying attention to the underlying resource composition of the cloud platform, thereby improving the user's task attention and the experience of task execution on the cloud platform; and in the process of generating the basic environment orchestration result, the available resources on the cloud platform are effectively integrated to improve the resource utilization of the cloud platform.

[0050] Figure 2 Schematic diagram of another basic environment orchestration method based on a cloud platform according to an embodiment of the present disclosure. This embodiment is a further refinement of the above technical solution. The technical solution in this embodiment can be combined with each optional solution in one or more of the above embodiments. Figure 2 As shown in the figure, the basic environment orchestration method based on the cloud platform includes the following:

[0051] S210, obtaining the functional requirement information and target network segment of the user for the cloud platform to issue the execution task.

[0052] S220: Decompose the execution task according to the functional requirement information to obtain multiple subtasks.

[0053] S230: Determine corresponding user individual requirements and task inherent requirements according to the execution attribute information of the subtask.

[0054] The execution attribute information is used to determine the execution process requirements and execution purpose of the subtask, including the user's individual needs, which are used to characterize the user's personal requirements for the execution process and execution purpose of the subtask, and the task-specific needs, which are used to characterize the indispensable features of the task execution and the features of the best operation realization. Exemplarily, the user's individual needs refer to the user's requirements for the type of machine that executes the subtask or the requirements for execution efficiency, etc., and the task-specific needs refer to the best operation plan for the subtask on the cloud platform, that is, which machine on the cloud platform is used to execute the subtask so that the execution efficiency of the subtask is the best.

[0055] Specifically, the content matching the subtask in the functional requirement information is used as the execution attribute information, and the execution attribute information is decomposed to obtain the user individual requirement and the task inherent requirement corresponding to the subtask.

[0056] S240: Determine the target platform resources corresponding to each subtask from the candidate platform resources of the cloud platform according to the individual needs of the user and the inherent needs of the task.

[0057] According to the user's individual needs and the inherent needs of the task, matching target platform resources are selected from the candidate platform resources. If there is a conflict between the user's individual needs and the inherent needs of the task for the same type of platform resources, the corresponding target platform resources are determined according to the user's individual needs.

[0058] Exemplarily, different types of computing cards are configured for subtasks of different task types according to the inherent requirements of each subtask, and different storage systems are used for subtasks with different requirements according to the user's individual needs of each subtask. For example, products such as parallel file systems are used to accelerate model training, and compliance security settings are performed for corresponding subtasks according to the user's individual needs of each subtask to prevent sensitive data leakage.

[0059] Based on the different types of candidate platform resources provided by the cloud platform, the underlying platform resources can be flexibly combined according to the different requirements of users and the tasks themselves. On the premise of meeting the task execution functions, costs can be controlled according to user requirements, and users are exempted from the operation and maintenance management of various candidate platform resources.

[0060] like Figure 3 The figure shows the architecture diagram of the execution of the basic environment orchestration method based on the cloud platform. The AI-Infra module is the main module for executing the basic environment orchestration method, which implements the cloud computing infrastructure, flexibly combines edge computing resources and central cloud computing resources, and realizes various functions for end users. End users train and deploy AI tasks through the business layer to realize various user-oriented functions, such as creating training clusters, queues, deploying inference services, etc., and call AI-Infra according to user needs, requiring the realization of resource combinations that meet user needs.

[0061] AI-Infra is responsible for producing and managing user clusters on the data plane of the platform according to the needs of the business layer, shielding the complexity of the basic environment orchestration from the business layer. According to the resource orchestration specifications submitted by users of the upper business layer, it automatically calls edge computing resources and central cloud computing resources to complete the construction of the computing power foundation.

[0062] Specifically, AI-Infra is located between the business layer and the infrastructure layer in the architectural layering, and automatically utilizes the capabilities of the infrastructure layer to generate resource orchestration that meets the needs of the business layer. The AI-Infra module is divided into two parts: AI-Infra-Service: used to provide functional requirement information to the upper business layer and generate basic environment orchestration specifications; AI-Infra-operator: used to automatically call the infrastructure API according to the basic environment orchestration specifications to complete the automatic organization of computing resources; wherein the infrastructure includes at least edge computing resources and central cloud computing resources. In addition, AI-Infra also has the function of scheduling components according to the functional requirement information provided by the business layer. The specific component scheduling function is not limited in the embodiments of the present invention.

[0063] In another optional implementation of this embodiment, the candidate platform resources include candidate computing resources, candidate storage resources, candidate network resources, and candidate component resources.

[0064] The cloud platform supports automated orchestration of multiple types of resources, including candidate computing resources, candidate storage resources, candidate network resources, and candidate component resources.

[0065] Among them, candidate computing resources include central cloud computing resources and third-party computing resources from third-party resource providers. Computing resources include virtual machines, bare metal and other types of GPU and CPU computing resources. Candidate storage resources include storage systems that support object storage, parallel file storage, etc. Candidate network resources include basic network facilities such as peer-to-peer connections, elastic network cards, dedicated lines, and routing tables. Candidate component resources include various components of the cloud native ecosystem to complete functional requirements such as AI scheduling, GPU virtualization, log collection, and observability. It also includes candidate cluster resources for connecting to the Kubernetes ecosystem and building a cloud native ecosystem consistent with the open source community.

[0066] This embodiment improves the integrity of the basic environment of the orchestration by realizing the automatic orchestration of various types of platform resources in the cloud platform, thereby improving the user experience.

[0067] In another optional implementation of this embodiment, S240 includes:

[0068] According to the individual needs of users and inherent requirements of tasks, the target computing resources corresponding to each subtask are determined from the candidate computing resources of the cloud platform;

[0069] According to the type of the target computing resource, the target network resource for interaction between subtasks is determined from the candidate network resources.

[0070] First, the target computing resources for executing the subtask are determined according to the user's individual needs and the inherent needs of the task corresponding to the subtask, and then the target network resources for the interconnection and interaction between the subtasks are determined according to the type of computing resources.

[0071] Furthermore, according to the individual needs of the user and the inherent needs of the task, the target computing resources corresponding to each subtask are determined from the candidate computing resources of the cloud platform, and the target storage resources corresponding to each subtask are determined from the candidate storage resources of the cloud platform. According to the type of the target computing resource and the type of the target storage resource, the target network resource for interaction between subtasks is determined from the candidate network resources. Exemplarily, the network connectivity and isolation solutions are automatically built according to the user's network connectivity requirements, computing power requirements, storage requirements, etc.

[0072] This embodiment first determines the target computing resources corresponding to each subtask, and then determines the target network resources corresponding to each subtask based on the target computing resources, thereby improving the accuracy of determining the target network resources and thus improving the accuracy of basic environment orchestration.

[0073] In another optional implementation of this embodiment, determining the target network resource for interaction between subtasks from the candidate network resources according to the type of the target computing resource includes:

[0074] If the target computing resource type of the subtask is a central cloud computing resource, then the elastic network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources; the intelligent network is used as the target network resource for communication and interaction with other subtasks that use edge computing resources; and peer-to-peer connection or elastic network card is used as the target network resource for communication and interaction with user resources;

[0075] If the target computing resource type of the subtask is edge computing resources, the service network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources.

[0076] The cluster heterogeneous computing power determined for the user's execution task includes various types of GPU resources, which are provided by the central cloud and third-party suppliers. The Kubernetes cluster computing control plane and the heterogeneous computing power are interconnected and interacted using elastic network cards (ENI). The central cloud and third-party computing power suppliers are previously connected using intelligent networks (dedicated lines, CSN). Third-party computing power suppliers can use service network cards to access the central cloud computing resources. GPU heterogeneous computing power and user VPCs are connected through peer connections or elastic network cards, and can access object storage, parallel file storage, etc. in the user VPC. At the same time, security groups, ACLs, etc. are automatically configured to ensure security while preserving the network connectivity of heterogeneous computing power and prevent possible information leakage. Figure 4 The figure shows the interconnection diagram of cluster heterogeneous computing resource pools. The cloud platform provides users with maintenance-free, exclusive resource pools, namely managed resource pools. Users can submit training tasks and deploy inference services without worrying about the composition of underlying resources. From the perspective of resource orchestration, it has achieved the following: automatic deployment of managed computing clusters and CCE clusters for users. The control plane of the computing cluster is hosted by the CCE cluster and deployed in the CCE service VPC; and automatic use of container networks for computing configuration.

[0077] For example, Figure 4 As shown, user A provides two VPCs to form the target network segment. A resource pool is deployed in each VPC. A central cloud cluster is deployed for resource pool 1. The central cloud cluster includes multiple nodes. Two nodes are used as examples in the figure. In each node, the corresponding task service of the user is executed through the user container. Each resource pool is managed by the corresponding CCEVPC, that is, the node in CCE VPC-1 is the management node that manages the computing nodes in resource pool 1. The primary network card ENI of the management node in CCEVPC-1 is used for internal communication between the nodes, the managed Master is used to execute the tasks corresponding to the management node, and the auxiliary network card ENI is used to establish the connection between the management node in CCE VPC-1 and all the computing nodes in the corresponding resource pool 1; similarly, a central cloud cluster and an edge cluster are deployed for resource pool 2. The central cloud cluster and the edge cluster in a resource pool are connected through a CSN dedicated line. The auxiliary network card ENI in CCE VPC-2 is used to establish the connection between the management node in CCE VPC-2 and all the computing nodes in the central cloud cluster and all the computing nodes in the edge cluster in the corresponding resource pool 2.

[0078] This embodiment determines the target network resources for interaction between subtasks according to the types of target computing resources corresponding to different subtasks, thereby improving the efficiency of network interaction between subtasks.

[0079] S250: Generate a basic environment orchestration result according to target platform resources and multiple subnets obtained by dividing the target network segment, and provide the basic environment orchestration result to the user for task execution.

[0080] The solution of this embodiment, by decomposing the execution task, determines the corresponding target platform resources as the basic environment for task execution according to the user's individual needs and the inherent needs of the task for each subtask, thereby improving the accuracy of determining the composition of the basic environment.

[0081] Figure 5It is a schematic diagram of another basic environment orchestration method based on a cloud platform according to an embodiment of the present disclosure. This embodiment further refines the target network segment division of the above-mentioned technical solution. The technical solution in this embodiment can be combined with various optional solutions in one or more of the above-mentioned embodiments.

[0082] like Figure 5 As shown in the figure, the basic environment orchestration method based on the cloud platform includes the following:

[0083] S310, obtaining the functional requirement information and target network segment of the user for the cloud platform to issue the execution task.

[0084] S320. Determine corresponding platform resource information from the cloud platform according to the functional requirement information.

[0085] S330: Based on the predetermined resource requirements of multiple network segments, divide the target network segment to obtain multiple subnets.

[0086] Specific objects that have a demand for using the network segment IP provided by the user during the task execution process on the cloud platform are counted in advance, and are summarized and divided according to the content of the specific objects to obtain multiple network segment resource requirements, that is, the network segment resource requirements are specific objects that have a demand for using the IP of the target network segment provided by the user.

[0087] Specifically, according to the network segment resource size information required by multiple network segment resource requirements, the target network segment provided by the user is divided, and a corresponding subnet of appropriate size is divided for each network segment resource requirement, so as to improve the efficiency and security of the user's private network access from the corresponding subnet for each network segment resource requirement. Exemplarily, the network segment resource size information required for each network segment resource requirement is determined according to the specific content of the network segment resource requirement, which is not limited here.

[0088] In another optional implementation of this embodiment, the predetermined multiple network segment resource demands include at least the demand of the computing node management node for network segments, the demand of the central cloud computing node for network segments, the demand of the user access portal address management for network segments, the demand of the edge computing node for network segments, the demand of the edge computing node for network segments, the demand of the central cloud public service access to the edge computing node for network segments, the demand of the container for network segments, and the demand of the container-to-container communication for network segments.

[0089] Among them, the computing node refers to the underlying computing power in the platform resource information determined for the user. For example, a computing node can correspond to a machine. Due to the large amount of tasks performed on the cloud platform, the platform resource information determined for the user includes multiple computing nodes. The management node of the computing node refers to the node that manages all computing nodes in the platform resource information. The management node can also be deployed on at least one machine in the platform resource information to manage the task distribution and scheduling of the computing node. The demand of the management node of the computing node for the network segment refers to the access demand of the management node to the target network segment provided by the user when managing the computing node, such as the access demand of the management node when acquiring data from the user's private network.

[0090] The central cloud computing node refers to the computing machine directly provided by the cloud platform. The demand of the central cloud computing node for the network segment refers to the access demand of the central cloud computing node in the cloud platform to the target network segment provided by the user during operation.

[0091] The user access entry address refers to the management address provided by the cloud platform to the user for the user to access the task. The demand for network segments in the user access entry address management refers to the access demand for the target network segment provided by the user generated by managing multiple access addresses generated during the execution of the task.

[0092] Edge computing nodes refer to computing machines provided by third parties other than the cloud platform. Since the types and number of computing machines on the cloud platform are limited, in order to improve the task execution efficiency of the cloud platform and reduce costs, the cloud platform will provide users with third-party computing machines, that is, to integrate the central cloud computing nodes and edge computing nodes of the cloud platform to increase the types and number of machines available for users to choose. The demand for network segments by edge computing nodes refers to the access requirements of the target network segments provided by users that edge computing nodes in the cloud platform need to use during operation.

[0093] Central cloud public services refer to public services provided by the public network where the central cloud is located. Central cloud computing nodes can directly access central cloud public services. Since edge computing nodes do not have a public network by default, some public services that edge computing nodes and container clusters rely on (such as yum / apt software sources, etc.) need to access the central cloud's public services through CSN (Cloud Society Network). Since the network segments used by public services have default internal routes (not visible to users) inside gateways such as VPC and CSN, the network segment routes explicitly added to the product will be invalid. Therefore, it is necessary to proxy the edge computing nodes' access to the central cloud public services through the proxy access method of the service network card or BLB in the central cloud.

[0094] A container is the smallest operation management unit created for task execution. Multiple containers form a computing node, and multiple computing nodes form a cluster. During task execution, the subtasks corresponding to the container also need to access data on the network segment provided by the user. During task execution, communication and interaction between containers also need to access user resources.

[0095] The embodiment of the present invention determines the resource requirements of multiple network segments according to the specific process of executing the task, thereby improving the accuracy of determining the resource requirements of the network segments, and further improving the accuracy of dividing the network segments.

[0096] In another optional implementation of this embodiment, based on a plurality of predetermined network segment resource requirements, the target network segment is divided to obtain a plurality of subnets, including:

[0097] Based on the demand of the management node of the computing node for the network segment, the first IP reservation quantity is determined from the target network segment according to the management cluster size and the preset available redundancy;

[0098] Based on the demand of the central cloud computing node for the network segment, the number of second IP reservations is determined according to the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool;

[0099] Based on the user access entry address management's demand for network segments, the number of third IP reservations is determined according to the business scenario of the user's task execution;

[0100] Based on the demand of edge computing nodes for network segments, the fourth IP reservation quantity is determined according to the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool;

[0101] Based on the demand for network segments to provide central cloud public service access to edge computing nodes, the number of fifth IP reservations is determined according to the proxy access method;

[0102] Based on the container's demand for network segments, the number of reserved sixth IP addresses is determined according to the maximum number of containers supported by a single cluster.

[0103] Based on the demand for network segments for communication between containers, the number of reserved seventh IP addresses is determined according to the network access volume between containers and user resources;

[0104] The target network segment is divided according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain multiple subnets.

[0105] The management cluster size refers to the number of computing nodes managed by the management node, and the preset available redundancy refers to the redundancy set for the access requirements of the management node to improve the access efficiency of the management node.

[0106] Specifically, the management cluster size is determined according to the number of computing nodes managed by the management node, for example, a mapping relationship between multiple computing node number intervals and candidate management cluster sizes is pre-established, and the corresponding IP reservation number is determined according to the management cluster size, for example, a mapping relationship between the candidate management cluster size and the IP reservation number is pre-established, and finally the first IP reservation number is determined according to the management cluster size, the corresponding IP reservation number and the IP reservation number corresponding to the preset available redundancy, that is, the management node can achieve efficient access to the user's private network through the first IP reservation number IPs in the target network segment. Exemplarily, consider evaluating the LB IP reservation number based on the Master LB (Load Balancer, LB)) high availability redundancy, evaluating the apiserver ENI (Elastic Network Interface) IP reservation number based on the Wanka hosting cluster size, and determining the first IP reservation number based on the LB IP reservation number and the apiserver ENI IP reservation number.

[0107] The maximum number of central cloud resource pools supported by a single user is determined based on the functional requirement information of the user's historical task execution, and the maximum number of nodes supported by a single central cloud resource pool is determined based on the attribute information set by the central cloud resource pool, and then the second IP reservation number is determined based on the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool. Exemplarily, the total number of central cloud nodes is determined based on the product of the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool, and the second IP reservation number is determined based on the mapping relationship between the predetermined central cloud node number interval and the IP reservation number. Exemplarily, during the task execution process, the subnet is divided into the central cloud resource pool granularity, and the peer connection routing is established to the subnet granularity. The resource pool perceived by the user is actually connected to the subnet, so the subnet is divided according to the maximum number of resource pools supported by a single user and the maximum number of nodes supported by a single resource pool.

[0108] The business scenario in which the user sends the execution task is determined according to the type of execution task. For example, the business scenarios include development machine, reasoning service, training task, etc. Specifically, a mapping relationship between the number of IP reservations corresponding to different business scenarios is established in advance, and the corresponding business scenario is determined according to the specific type of the execution task sent by the user, and then the third IP reservation number corresponding to the business scenario is determined according to the mapping relationship. Exemplarily, the third IP reservation number is estimated based on the number of independent BLBs that may be used in the business scenario (development machine, reasoning service, training task TensorBoard), while considering that some scenarios may have BLB multiplexing (single instance multiple ports).

[0109] The maximum number of edge resource pools supported by a single user is determined based on the usage information of the third-party computing nodes of the cloud platform, and the maximum number of nodes supported by a single edge resource pool is determined based on the attribute information set by the edge resource pool, and then the fourth IP reservation number is determined based on the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool. Exemplarily, the total number of edge nodes is determined based on the product of the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool, and the fourth IP reservation number is determined based on the mapping relationship between the predetermined edge node number interval and the IP reservation number. Exemplarily, the edge VPC (VirtualPrivate Cloud, VPC) and the central VPC are in a large network segment, and appear to the user to be the same network segment. The edge VPC network segment size and subnet division are determined based on the number of single-tenant edge resource pools and resource pool size that are planned to be supported.

[0110] Determine the proxy access method of the edge node to the central cloud public service, and determine the fifth IP reserved number based on the predetermined mapping relationship between different proxy access methods and IP reserved numbers.

[0111] The maximum number of containers supported by a single cluster is determined based on the attribute information of the central cloud, and then the sixth IP reservation number is determined based on the mapping relationship between the predetermined number of containers and the number of IP reservations. Exemplarily, the container network uses an Overlay network to avoid conflicts in accessing the user VPC network through containers. Therefore, a special network segment is also required to divide the container subnet according to the maximum number of containers supported by a single cluster.

[0112] During the task execution process, communication and interaction between containers also need to access user resources. Therefore, when the container accesses user resources through the user's target network segment, the number of reserved seventh IP addresses is determined based on the network access volume, where the network access volume between the container and the user's resources can be determined based on historical access data. Set a separate network segment for communication between containers to avoid Service unavailability due to conflicts with the user's VPC network.

[0113] According to the first IP reservation quantity, a corresponding subnet is divided from the target network as the hosted cluster Master subnet, which is used to provide IP access support for the Master LB and apiserver ENI. According to the second IP reservation quantity, a corresponding subnet seat resource pool subnet is divided from the target network to provide IP access support for the node ENI in the central cloud. According to the third IP reservation quantity, a corresponding subnet is divided from the target network as the LB Service subnet for BLB IP access support. According to the fourth IP reservation quantity, a corresponding subnet is divided from the target network as the edge VPC subnet to provide IP access support for edge nodes. According to the fifth IP reservation quantity, a corresponding subnet is divided from the target network as a public service subnet to provide IP access support for edge nodes to access the central cloud public services. According to the sixth IP reservation quantity, a corresponding subnet is divided from the target network as a container subnet, and according to the seventh IP reservation quantity, a corresponding subnet is divided from the target network as a cluster subnet for IP address access for container communication within the cluster.

[0114] like Figure 6 The target network segment division diagram is shown. The target network segment of user A includes three VPC network segments: VPC-1, VPC-2 and VPC-3. According to the needs of users, corresponding resource pools are set in each VPC network segment. The cloud platform includes platform resource information determined for users. The three VPC network segments are divided according to different network segment resource requirements. Among them, the VPC-1 network segment is divided to the first central cloud VPC. The first central cloud VPC includes node group 1, node group system and management node. Among them, the demand for access to the user's private network in the first central cloud VPC is realized through the VPC-1 network segment. Further, the VPC-1 network segment can be further divided according to the respective needs of the node group and the management node. Similarly, the VPC-2 network segment is divided to the second central cloud VPC and the second edge VPC. Among them, the demand for access to the user's private network in the second central cloud VPC and the second edge VPC is realized through the VPC-2 network segment. Further, the VPC-2 network segment can be further divided according to the respective needs of the node group and the management node. The VPC-3 network segment is divided into the third central cloud VPC and the third edge VPC, wherein the demand for access to the user's private network in the third central cloud VPC and the third edge VPC is realized through the VPC-3 network segment. Furthermore, the VPC-3 network segment can be further divided according to the respective needs of the node group and the management node.

[0115] If a central cloud node in the first central cloud VPC, the second central cloud VPC, and the third central cloud VPC is offline or missing, a backup central cloud node is determined from the backup pool through the node change VPC; similarly, if an edge node in the second edge VPC and the third edge VPC is offline or missing, a backup edge node is determined from the backup pool through the node change VPC.

[0116] In another optional implementation of the present embodiment, before dividing the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number, and the seventh IP reserved number to obtain multiple subnets, the method further includes:

[0117] Determine a comparison result of the estimated sum of IPs of the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity, and the seventh IP reserved quantity with the sum of target IPs included in the target network segment;

[0118] If the total number of target IPs is greater than the total number of expected IPs, the number of remaining IPs is determined based on the total number of target IPs and the total number of expected IPs;

[0119] The first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity are updated according to the remaining IP quantity.

[0120] In order to avoid redundant waste of IPs in the target network segment provided by users, after determining the corresponding number of reserved IPs according to different network segment resource requirements, since the currently determined number of reserved IPs is the minimum access requirement for different network segment resource requirements, after determining the estimated total number of IPs based on the minimum access requirement, determine whether there are any remaining IPs in the target network segment based on the estimated total number of IPs.

[0121] Specifically, the sum of the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number is taken as the expected IP sum. The comparison result of the expected IP sum and the target IP sum included in the target network segment is determined. If the expected IP sum is the same as the target IP sum, network segment planning is performed according to the currently determined first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain multiple subnets.

[0122] If the total number of target IPs is greater than the total number of expected IPs, it means that there are remaining IPs in the target network segment provided by the user, and the remaining IPs are allocated to different network segment resource requirements, that is, the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity, and the seventh IP reserved quantity are updated. Exemplarily, the priorities of different network segment resource requirements are determined in advance, and different IP reserved quantities are updated from high to low priority; or the remaining IPs are divided according to the proportion of the IP reserved quantities currently corresponding to different network resource requirements, and the remaining IPs are allocated to different IP reserved quantities according to the proportion.

[0123] If the total number of target IPs is less than the total number of expected IPs, it means that there is a problem of insufficient target network segments when dividing network segments according to the current plan, so the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number are reduced. Exemplarily, the priorities of different network segment resource requirements are determined in advance, and the different IP reserved numbers are reduced from low to high priority; or the remaining IPs are divided according to the proportion of the IP reserved numbers currently corresponding to different network resource requirements, and the IP reserved numbers corresponding to different network segment resource requirements are reduced according to the proportion, so that the total number of reduced IPs is the difference between the expected IP sum and the target IP sum.

[0124] This embodiment makes full use of the target network segment provided by the user and avoids wasting the resources provided by the user by determining the size comparison result between the target IP sum and the expected IP sum.

[0125] S340: Generate a basic environment orchestration result according to the platform resource information and multiple subnets, and provide the basic environment orchestration result to the user for task execution.

[0126] The solution of this embodiment divides corresponding subnets for multiple network segment resource requirements from different perspectives, and implements access to the target network segments provided by users through different subnets for different network resource requirements, thereby improving the efficiency of access to the target network segments, avoiding conflicts in access to the target network segments due to different requirements during task execution, and improving the accuracy of division of the target network segments.

[0127] Figure 7 is a schematic diagram of the structure of a basic environment arrangement device based on a cloud platform according to an embodiment of the present disclosure, and the device can execute the basic environment arrangement method based on a cloud platform involved in any embodiment of the present disclosure; Figure 7 The basic environment orchestration device 700 based on the cloud platform includes: a function requirement acquisition module 710, a platform resource determination module 720 and an environment orchestration module 730.

[0128] Function requirement acquisition module 710, used to obtain the function requirement information and target network segment of the user for the cloud platform to issue the execution task;

[0129] The platform resource determination module 720 is used to determine corresponding platform resource information from the cloud platform according to the functional requirement information;

[0130] The environment orchestration module 730 is used to generate a basic environment orchestration result according to the platform resource information and multiple subnets obtained by dividing the target network segment, and provide the basic environment orchestration result to the user for task execution.

[0131] The solution of this embodiment determines the corresponding platform resource information through the functional requirement information of the task to be executed, and divides the target network segment provided by the user, and then generates the basic environment orchestration result of the task execution according to the platform resource information and the subnet obtained by the division, so that the user can execute the task that meets the user's needs without paying attention to the underlying resource composition of the cloud platform, thereby improving the user's task attention and the experience of task execution on the cloud platform; and in the process of generating the basic environment orchestration result, the available resources on the cloud platform are effectively integrated to improve the resource utilization of the cloud platform.

[0132] In an optional implementation of this embodiment, the platform resource determination module includes:

[0133] A subtask decomposition submodule is used to decompose the execution task according to the functional requirement information to obtain multiple subtasks;

[0134] The target platform resource determination submodule is used to determine the corresponding target platform resources from the candidate platform resources of the cloud platform for each subtask.

[0135] In an optional implementation of this embodiment, the target platform resource determination submodule includes:

[0136] A demand determination unit, used to determine corresponding user individual demands and task inherent demands according to the execution attribute information of the subtask;

[0137] The target platform resource determination unit is used to determine the target platform resource corresponding to each of the subtasks from the candidate platform resources of the cloud platform according to the user's individual needs and the inherent needs of the task.

[0138] In an optional implementation of this embodiment, the candidate platform resources include candidate computing resources, candidate storage resources, candidate network resources, and candidate component resources.

[0139] In an optional implementation of this embodiment, the target platform resource determination unit is specifically configured to:

[0140] A computing resource determination subunit, used to determine a target computing resource corresponding to each of the subtasks from candidate computing resources of the cloud platform according to the individual needs of the user and the inherent needs of the task;

[0141] The network resource determination subunit is used to determine the target network resource for interaction between the subtasks from the candidate network resources according to the type of the target computing resource.

[0142] In an optional implementation of this embodiment, the network resource determination subunit includes:

[0143] If the target computing resource type of the subtask is a central cloud computing resource, an elastic network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources; an intelligent network is used as the target network resource for communication and interaction with other subtasks that use edge computing resources; and a peer-to-peer connection or an elastic network card is used as the target network resource for communication and interaction with user resources;

[0144] If the target computing resource type of the subtask is an edge computing resource, a service network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources.

[0145] In an optional implementation of this embodiment, the device further includes a subnet division module, which is used, before generating a basic environment orchestration result according to the platform resource information and the multiple subnets obtained by dividing the target network segment, to:

[0146] Based on a plurality of predetermined network segment resource requirements, the target network segment is divided to obtain a plurality of subnets.

[0147] In an optional implementation of this embodiment, the predetermined multiple network segment resource demands include at least the demand of the computing node management node for network segments, the demand of the central cloud computing node for network segments, the demand of the user access portal address management for network segments, the demand of the edge computing node for network segments, the demand of the edge computing node for network segments, the demand of the central cloud public service access to the edge computing node for network segments, the demand of the container for network segments, and the demand of the container-to-container communication for network segments.

[0148] In an optional implementation of this embodiment, the subnet division module includes:

[0149] A first demand division submodule, used to determine a first IP reserved quantity from the target network segment based on the demand of the management node of the computing node for the network segment according to the management cluster size and preset available redundancy, and determine a first divided subnet according to the first IP reserved quantity;

[0150] A second demand division submodule is used to determine the second IP reservation quantity based on the demand of the central cloud computing node for the network segment, according to the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool, and determine the second divided subnet according to the second IP reservation quantity;

[0151] A third demand division submodule is used to manage the demand for network segments based on the user access entry address, determine the third IP reserved quantity according to the business scenario of the user issuing the execution task, and determine the third divided subnet according to the third IP reserved quantity;

[0152] A fourth demand division submodule is used to determine a fourth IP reserved quantity based on the demand of the edge computing node for the network segment, according to the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool, and determine a fourth divided subnet according to the fourth IP reserved quantity;

[0153] A fifth demand division submodule, configured to determine a fifth IP reserved quantity according to a proxy access mode based on the demand for network segments provided to edge computing nodes for providing central cloud public service access, and determine a fifth divided subnet according to the fifth IP reserved quantity;

[0154] A sixth demand division submodule, used to determine a sixth IP reserved number based on the container's demand for the network segment and the maximum number of containers supported by a single cluster, and determine a sixth divided subnet based on the sixth IP reserved number;

[0155] A seventh demand division submodule, used to determine the seventh IP reserved quantity based on the demand for network segments for communication between containers and network access volume between containers and user resources, and determine the seventh divided subnet based on the seventh IP reserved quantity;

[0156] The target network segment division submodule is used to divide the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain multiple subnets.

[0157] In an optional implementation of this embodiment, the device further includes a remaining network segment division submodule, which is used to divide the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number, and the seventh IP reserved number to obtain a plurality of subnets, including:

[0158] An IP comparison unit is used to determine a comparison result between the estimated IP sum of the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity and the target IP sum included in the target network segment;

[0159] A remaining IP determination unit, configured to determine the number of remaining IPs according to the target IP sum and the expected IP sum if the target IP sum is greater than the expected IP sum;

[0160] The remaining IP division unit is used to update the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity according to the remaining IP quantity.

[0161] The above-mentioned basic environment orchestration device based on the cloud platform can execute the basic environment orchestration method based on the cloud platform provided by any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not described in detail in this embodiment, please refer to the basic environment orchestration method based on the cloud platform provided by any embodiment of the present disclosure.

[0162] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0163] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0164] Figure 8A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0165] like Figure 8 As shown, the device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0166] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0167] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as a basic environment orchestration method based on a cloud platform. For example, in some embodiments, the basic environment orchestration method based on a cloud platform may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the basic environment orchestration method based on a cloud platform described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured in any other appropriate manner (eg, by means of firmware) to execute the method based on the basic environment orchestration of the cloud platform.

[0168] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0169] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0170] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0171] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0172] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0173] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0174] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0175] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A basic environment orchestration method based on a cloud platform, comprising: Obtain the functional requirement information and target network segment of the user for the cloud platform to issue execution tasks; According to the functional requirement information, determining corresponding platform resource information from the cloud platform; A basic environment orchestration result is generated according to the platform resource information and a plurality of subnets obtained by dividing the target network segment, and the basic environment orchestration result is provided to a user for task execution.

2. The method according to claim 1, wherein: Determining corresponding platform resource information from the cloud platform according to the functional requirement information includes: Decomposing the execution task according to the functional requirement information to obtain multiple subtasks; Determine the corresponding target platform resources from the candidate platform resources of the cloud platform for each subtask respectively.

3. The method according to claim 2, wherein: The step of respectively determining the corresponding target platform resources from the candidate platform resources of the cloud platform for each subtask includes: Determining corresponding user individual requirements and task inherent requirements according to the execution attribute information of the subtask; According to the individual needs of the user and the inherent needs of the task, a target platform resource corresponding to each of the subtasks is determined from the candidate platform resources of the cloud platform.

4. The method according to claim 2 or 3, wherein: The candidate platform resources include candidate computing resources, candidate storage resources, candidate network resources and candidate component resources.

5. The method according to claim 4, wherein: The determining, according to the user's individual needs and the inherent needs of the task, a target platform resource corresponding to each of the subtasks from the candidate platform resources of the cloud platform includes: Determine the target computing resource corresponding to each of the subtasks from the candidate computing resources of the cloud platform according to the individual needs of the user and the inherent needs of the task; According to the type of the target computing resource, a target network resource for interaction between the subtasks is determined from candidate network resources.

6. The method according to claim 5, wherein: The step of determining the target network resource for interaction between the subtasks from candidate network resources according to the type of the target computing resource includes: If the target computing resource type of the subtask is a central cloud computing resource, an elastic network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources; an intelligent network is used as the target network resource for communication and interaction with other subtasks that use edge computing resources; and a peer-to-peer connection or an elastic network card is used as the target network resource for communication and interaction with user resources; If the target computing resource type of the subtask is an edge computing resource, a service network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources.

7. The method according to claim 1, wherein: Before generating a basic environment orchestration result according to the platform resource information and the multiple subnets obtained by dividing the target network segment, the method further includes: Based on a plurality of predetermined network segment resource requirements, the target network segment is divided to obtain a plurality of subnets.

8. The method according to claim 7, wherein: The predetermined multiple network segment resource demands include at least the demands of the computing node management node for network segments, the demands of the central cloud computing node for network segments, the demands of the user access portal address management for network segments, the demands of the edge computing node for network segments, the demands of the edge computing node for network segments, the demands of the container for network segments, and the demands of the container-to-container communication for network segments.

9. The method according to claim 8, wherein: The target network segment is divided based on the predetermined multiple network segment resource requirements to obtain multiple subnets, including: Based on the demand of the management node of the computing node for the network segment, determining the first IP reservation quantity from the target network segment according to the management cluster size and preset available redundancy; Based on the demand of the central cloud computing node for the network segment, the number of second IP reservations is determined according to the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool; Based on the user access entry address management's demand for network segments, the number of third IP reservations is determined according to the business scenario of the user's task execution; Based on the demand of edge computing nodes for network segments, the fourth IP reservation quantity is determined according to the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool; Based on the demand for network segments to provide central cloud public service access to edge computing nodes, the number of fifth IP reservations is determined according to the proxy access method; Based on the container's demand for network segments, the number of reserved sixth IP addresses is determined according to the maximum number of containers supported by a single cluster. Based on the demand for network segments for communication between containers, the number of reserved seventh IP addresses is determined according to the network access volume between containers and user resources; The target network segment is divided according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain multiple subnets.

10. The method according to claim 9, before dividing the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain a plurality of subnets, the method further comprises: Determine a comparison result of the estimated sum of the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity, and the seventh IP reserved quantity with the sum of the target IPs included in the target network segment; If the target IP sum is greater than the expected IP sum, the remaining IP sum is determined based on the target IP sum and the expected IP sum; The first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity are updated according to the remaining IP quantity.

11. A basic environment arrangement device based on a cloud platform, comprising: Functional requirement acquisition module, used to obtain the functional requirement information and target network segment of the user for the cloud platform to issue execution tasks; A platform resource determination module, used to determine corresponding platform resource information from the cloud platform according to the functional requirement information; The environment orchestration module is used to generate a basic environment orchestration result according to the platform resource information and a plurality of subnets obtained by dividing the target network segment, and provide the basic environment orchestration result to the user for task execution.

12. The device according to claim 11, wherein The platform resource determination module includes: A subtask decomposition submodule is used to decompose the execution task according to the functional requirement information to obtain multiple subtasks; The target platform resource determination submodule is used to determine the corresponding target platform resource from the candidate platform resources of the cloud platform for each subtask.

13. The device according to claim 12, wherein: The target platform resource determination submodule includes: A demand determination unit, used to determine corresponding user individual demands and task inherent demands according to the execution attribute information of the subtask; The target platform resource determination unit is used to determine the target platform resource corresponding to each of the subtasks from the candidate platform resources of the cloud platform according to the user's individual needs and the inherent needs of the task.

14. The device according to claim 12 or 13, wherein: The candidate platform resources include candidate computing resources, candidate storage resources, candidate network resources and candidate component resources.

15. The device according to claim 14, wherein: The target platform resource determination unit is specifically used to: A computing resource determination subunit, used to determine a target computing resource corresponding to each of the subtasks from candidate computing resources of the cloud platform according to the individual needs of the user and the inherent needs of the task; The network resource determination subunit is used to determine the target network resource for interaction between the subtasks from the candidate network resources according to the type of the target computing resource.

16. The device according to claim 15, wherein: The network resource determination subunit includes: If the target computing resource type of the subtask is a central cloud computing resource, an elastic network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources; an intelligent network is used as the target network resource for communication and interaction with other subtasks that use edge computing resources; and a peer-to-peer connection or an elastic network card is used as the target network resource for communication and interaction with user resources; If the target computing resource type of the subtask is an edge computing resource, a service network card is used as the target network resource for communication and interaction with other subtasks that use central cloud computing resources.

17. The device according to claim 11, wherein: The device further includes a subnet division module, which is used, before generating a basic environment orchestration result according to the platform resource information and the multiple subnets obtained by dividing the target network segment, to: Based on a plurality of predetermined network segment resource requirements, the target network segment is divided to obtain a plurality of subnets.

18. The device according to claim 17, wherein: The predetermined multiple network segment resource demands include at least the demands of the computing node management node for network segments, the demands of the central cloud computing node for network segments, the demands of the user access portal address management for network segments, the demands of the edge computing node for network segments, the demands of the edge computing node for network segments, the demands of the container for network segments, and the demands of the container-to-container communication for network segments.

19. The device according to claim 18, wherein: The subnet division module comprises: A first demand division submodule, used to determine a first IP reserved quantity from the target network segment based on the demand of the management node of the computing node for the network segment according to the management cluster size and preset available redundancy, and determine a first divided subnet according to the first IP reserved quantity; A second demand division submodule is used to determine the second IP reservation quantity based on the demand of the central cloud computing node for the network segment, according to the maximum number of central cloud resource pools supported by a single user and the maximum number of nodes supported by a single central cloud resource pool, and determine the second divided subnet according to the second IP reservation quantity; A third demand division submodule is used to manage the demand for network segments based on the user access entry address, determine the third IP reserved quantity according to the business scenario of the user issuing the execution task, and determine the third divided subnet according to the third IP reserved quantity; A fourth demand division submodule is used to determine a fourth IP reserved quantity based on the demand of the edge computing node for the network segment, according to the maximum number of edge resource pools supported by a single user and the maximum number of nodes supported by a single edge resource pool, and determine a fourth divided subnet according to the fourth IP reserved quantity; A fifth demand division submodule, configured to determine a fifth IP reserved quantity according to a proxy access mode based on the demand for network segments provided to edge computing nodes for providing central cloud public service access, and determine a fifth divided subnet according to the fifth IP reserved quantity; A sixth demand division submodule, used to determine a sixth IP reserved number based on the container's demand for the network segment and the maximum number of containers supported by a single cluster, and determine a sixth divided subnet based on the sixth IP reserved number; A seventh demand division submodule, used to determine the seventh IP reserved quantity based on the demand for network segments for communication between containers and network access volume between containers and user resources, and determine the seventh divided subnet based on the seventh IP reserved quantity; The target network segment division submodule is used to divide the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain multiple subnets.

20. The device according to claim 19, wherein The device also includes a remaining network segment division submodule, which is used for dividing the target network segment according to the first IP reserved number, the second IP reserved number, the third IP reserved number, the fourth IP reserved number, the fifth IP reserved number, the sixth IP reserved number and the seventh IP reserved number to obtain a plurality of subnets, including: An IP comparison unit is used to determine a comparison result between the estimated IP sum of the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity and the target IP sum included in the target network segment; A remaining IP determination unit, configured to determine the number of remaining IPs according to the target IP sum and the expected IP sum if the target IP sum is greater than the expected IP sum; The remaining IP division unit is used to update the first IP reserved quantity, the second IP reserved quantity, the third IP reserved quantity, the fourth IP reserved quantity, the fifth IP reserved quantity, the sixth IP reserved quantity and the seventh IP reserved quantity according to the remaining IP quantity.

21. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

22. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-10.

23. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1-10 are implemented.

Images