Methods, apparatus, and systems for configuring serverless resources

Abstract

This invention provides a method, apparatus, and system for configuring serverless resources, relating to the field of resource configuration, and can provide users with workload deployment solutions that meet their needs. Based on this invention, the system can provide configuration items that allow users to define one or more attribute parameters among scaling and / or scheduling workloads, the type of computing resources used for the workload, or the workload's tolerance to interruptions. This supports users in determining resources based on their actual workload and preferences, achieving rational resource allocation. Furthermore, the system can estimate and display the cost of the required resources based on the user's attribute configuration, intuitively informing the user of the costs they will subsequently incur. This allows users to decide whether to accept the workload deployment solution after receiving payment information.

Description

Technical Field

[0001] This invention relates to the field of resource allocation, and more particularly to a method, apparatus, and system for configuring serverless resources. Background Technology

[0002] When users deploy their workloads (e.g., containers, non-containerized workloads, microservices, and batch applications) on a computing platform, they first purchase infrastructure resources such as virtual machines (VMs) or Kubernetes clusters, and then deploy their workloads on those resources. This approach provides users with high configuration and scaling flexibility. However, in this method, users ultimately pay for and manage all the overhead associated with the aforementioned infrastructure resources.

[0003] To reduce costs for users, they can deploy applications and workloads on serverless computing platforms. Serverless computing platforms allow users to deploy their applications / workloads without configuring or managing servers and other infrastructure resources. This way, users only pay for the resources (i.e., dedicated resources) used by their applications / workloads on the serverless computing platform, thus reducing the cost of managing dedicated resources.

[0004] However, when users deploy their applications / workloads on these serverless computing platforms, most of these platforms offer very limited configurability within their systems. Summary of the Invention

[0005] This invention provides a method, apparatus, and system for configuring serverless resources, which can support users to configure or adjust resources in detail and flexibly according to the resource requirements and preferences of actual workloads. At the same time, it allows users to intuitively understand the impact of different resource parameters on costs, thereby formulating a scientific, reliable workload deployment plan that meets user needs.

[0006] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions.

[0007] In a first aspect, a resource allocation method is provided. The method is executed in a serverless system and includes: receiving a first request including first parameters, wherein the first parameters include a first set of attribute parameters comprising one or more of scaling and / or scheduling a workload to be deployed, a computing resource type of the workload, or the workload's tolerance to interruption; estimating a first cost value based on the first parameters; outputting the first cost value; and, upon receiving first feedback on the first cost value, configuring resources for deploying the workload based on the first parameters.

[0008] For example, the workload to be deployed may include, but is not limited to, containerized application workloads, non-containerized application workloads, microservice workloads, and batch application workloads. This invention does not specifically limit the workload, and the workload may vary depending on the specific use case and business authorization conditions.

[0009] For example, the first feedback on the first cost value can determine the deployment of workloads based on the resources used for the first parameter corresponding to the first cost value.

[0010] In the solution provided in the first aspect above, the serverless system can offer configuration options, allowing users to define one or more attribute parameters such as the speed of scaling and / or scheduling workloads, the type of computing resources used for the workload, or the workload's tolerance to interruptions. This enables users to configure or adjust resources in detail and flexibly according to the actual workload and their own preferences, achieving reasonable resource allocation. Furthermore, the serverless system can estimate and display the cost of the required resources based on the user's attribute configuration, intuitively informing the user of the costs they will subsequently incur. This allows users to decide whether to accept the workload deployment plan after receiving payment information. In this way, the serverless system can tailor a scientific, reliable workload deployment plan that meets the user's specific needs.

[0011] In one possible implementation, the method further includes: before receiving the first request, in response to a template request, outputting a resource configuration template including a second parameter and a second cost value, wherein the first parameter is a parameter obtained by adjusting the second parameter, and the second parameter includes a set of second attribute parameters comprising one or more of the following: the speed at which the workload is scaled and / or scheduled, the type of computing resources used by the workload, or the workload's tolerance to interruptions. In this way, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates). Workloads can be deployed based on workload profiles corresponding to templates selected directly by the user from existing templates, simplifying user operations and reducing computation on the serverless system.

[0012] In another possible implementation, the method further includes: outputting first cost recommendation information, wherein the first cost recommendation information is used to indicate the impact of the difference between the first parameter and the second parameter on the cost of resources required to deploy the workload. This allows users to easily and intuitively understand the impact of different parameter sets on costs, enabling them to tailor a more scientific, reliable, and user-friendly workload deployment plan based on their own circumstances, thereby improving the user experience.

[0013] In another possible implementation, the attribute parameter for scaling and / or scheduling the workload includes the proportion of the workload to be deployed on at least one type of node among hot pool nodes, cold pool nodes, or instant pool nodes, wherein scaling and / or scheduling the hot pool nodes is faster than scaling and / or scheduling the cold pool nodes, and scaling and / or scheduling the cold pool nodes is faster than scaling and / or scheduling the instant pool nodes. In this way, the serverless system can support users in selecting different types of node resources based on the speed requirements of scaling and / or scheduling the workload, thereby developing a more scientific, reliable, and user-responsive workload deployment plan.

[0014] In another possible implementation, the attribute parameters of the computing resource type of the workload include at least one of input / output (I / O) intensive resources, central processing unit (CPU) intensive resources, or memory intensive resources. In this way, the serverless system can allow users to select different types of computing resources based on the specific needs of the workload in one or more aspects such as I / O performance, CPU performance, or memory performance, thereby developing a more scientific, reliable, and user-responsive workload deployment plan.

[0015] In another possible implementation, the workload's tolerance to interruption includes the following attribute parameters: the proportion of the workload to be deployed on on-demand instances, and / or the proportion of the workload to be deployed on spot instances. This allows serverless systems to support users in selecting instances with different levels of reliability based on their varying tolerance for interruption, thereby enabling more scientific, reliable, and user-responsive workload deployment solutions.

[0016] In another possible implementation, the method further includes: estimating a third cost value in response to second feedback on the first cost value, wherein the second feedback includes a third parameter, the third parameter including a set of third attribute parameters comprising one or more of the following: the speed at which the workload is scaled and / or scheduled, the type of computing resources used by the workload, or the workload's tolerance to interruptions; outputting the third cost value; and configuring resources for deploying the workload according to the third parameters after receiving third feedback on the third cost value. In this way, if a user is not satisfied with the first cost value, the serverless system allows the user to adjust attributes to obtain a workload deployment scheme that meets the user's cost expectations, thereby obtaining an alternative workload deployment scheme that satisfies the user's needs and preferences.

[0017] In another possible implementation, the method further includes: outputting second cost recommendation information, wherein the second cost recommendation information is used to indicate the impact of the difference between the first parameter and the third parameter on the cost of resources required to deploy the workload. This allows users to easily and intuitively understand the impact of different parameter sets on costs, enabling them to tailor a more scientific, reliable, and user-friendly workload deployment plan based on their own circumstances, thereby improving the user experience.

[0018] In a second aspect, a resource configuration method is provided. The method is executed in a terminal and includes: in response to a first input by a user on a first interface of the terminal, sending a first request to a serverless system including first parameters, wherein the first parameters include a first set of attribute parameters comprising one or more of scaling and / or scheduling a workload to be deployed, a computing resource type of the workload, or the workload's tolerance to interruption; receiving from the serverless system a first cost value of resources estimated according to the first parameters for deploying the workload; displaying a second interface including the first cost value; and in response to a second input by the user on the second interface, sending first feedback to the serverless system, wherein the first feedback is used to instruct the configuration of the resources for deploying the workload according to the first parameters.

[0019] For example, the first input is an operation on one or more attribute parameters on the first interface to configure or adjust the speed of scaling and / or scheduling of the workload, the computing resource type of the workload, or the workload's tolerance to interruptions.

[0020] In the solution provided in the second aspect above, the serverless system can provide configuration options through a terminal, allowing users to define one or more attribute parameters such as the speed of scaling and / or scheduling workloads, the type of computing resources used for the workload, or the workload's tolerance to interruptions. This supports users in making detailed and flexible configurations or adjustments to resources based on the actual workload and their own preferences, achieving reasonable resource allocation. Furthermore, the serverless system can estimate and display the cost of the required resources through the terminal based on the user's attribute configuration, intuitively informing the user of the costs to be paid. This allows users to decide whether to accept the workload deployment plan after knowing the payment information. In this way, the serverless system can develop a scientific, reliable workload deployment plan that meets the user's needs, taking into account their specific circumstances.

[0021] In one possible implementation, the method further includes: sending a template request to the serverless system in response to a third input by the user on a third interface, prior to sending a first request including the first parameter to the serverless system in response to a first input by the user on a first interface of the terminal; receiving a first resource configuration template from the serverless system including a second parameter and a second cost value, wherein the second parameter includes a set of second attribute parameters of one or more of scaling and / or scheduling the workload, the computing resource type of the workload, or the workload's tolerance to interruptions, and the second cost value is a cost value of resources estimated based on the second parameter for deploying the workload; and displaying the first interface including the second parameter and the second cost value, wherein the first parameter is a parameter obtained by adjusting the second parameter. In this way, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates) through the terminal. The workload can be deployed according to a workload profile corresponding to a template selected directly by the user from existing templates, simplifying user operation and reducing computation on the serverless system.

[0022] In another possible implementation, the second interface further includes a first view and a second view, wherein the first view is used to display the first parameter in geometric form, and the second view is used to display the second parameter in geometric form. This enhances the visualization of the parameter, thereby improving the user experience.

[0023] In another possible implementation, the method further includes: receiving first cost recommendation information from the server, wherein the first cost recommendation information is used to indicate the impact of the difference between the first parameter and the second parameter on the cost of resources required to deploy the workload, and the second interface also includes the first cost recommendation information. This allows users to easily and intuitively understand the impact of different parameter sets on costs, enabling them to combine their own circumstances with the serverless system to develop a more scientific, reliable, and user-friendly workload deployment plan, thereby improving the user experience.

[0024] In another possible implementation, the method further includes: in response to a fourth input from the user on the second interface, sending second feedback to the serverless system including a third parameter, wherein the third parameter is a parameter obtained by adjusting the first parameter, the third parameter including a set of third attribute parameters of one or more of scaling and / or scheduling the workload, the type of computing resources of the workload, and the workload's tolerance to interruptions, the second feedback being used to indicate a third cost value estimated based on the third parameter for deploying the workload; receiving the third cost value from the serverless system; and displaying a fourth interface including the third cost value. In this way, if the user is not satisfied with the first cost value, the serverless system allows the user to adjust attributes to obtain a workload deployment scheme that meets the user's cost expectations, thereby obtaining an alternative workload deployment scheme that meets the user's needs and preferences.

[0025] In another possible implementation, the fourth interface further includes a first view and a third view, wherein the first view is used to display the first parameter in geometric form, and the third view is used to display the third parameter in geometric form. This enhances the visualization of the parameter, thereby improving the user experience.

[0026] In another possible implementation, the method further includes: receiving second cost recommendation information from the serverless system, wherein the second cost recommendation information is used to indicate the impact of the difference between the first parameter and the third parameter on the cost of resources required to deploy the workload, and the fourth interface also includes the second cost recommendation information. This allows users to easily and intuitively understand the impact of different parameter sets on costs, enabling them to combine their own circumstances with the serverless system to develop a more scientific, reliable, and user-friendly workload deployment plan, thereby improving the user experience.

[0027] In another possible implementation, the method further includes receiving the fourth input before sending the second feedback to the serverless system. Receiving the fourth input includes receiving adjustments to the first view, or receiving attribute parameters input on the second interface, such as scaling and / or scheduling the speed of the workload, the computing resource type of the workload, or the workload's tolerance to interruptions. This allows users to configure or adjust attributes in multiple ways, thereby improving user operability.

[0028] In another possible implementation, the method further includes: in response to a fifth input from the user on the second interface, receiving from the serverless system a second resource configuration template including a fourth parameter and a fourth cost value, wherein the fourth parameter includes a set of fourth attribute parameters of one or more of the speed of scaling and / or scheduling the workload, the type of computing resources of the workload, or the tolerance of the workload to interruption, and the fourth cost value is a cost value of resources estimated based on the fourth parameter for deploying the workload; displaying a fifth interface including a first view and a fourth view, wherein the first view is used to display the first parameter geometrically, and the fourth view is used to display the fourth parameter geometrically. In this way, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates) via a terminal. The workload can be deployed according to a workload profile corresponding to a template selected directly by the user from existing templates, simplifying user operation and reducing computation on the serverless system.

[0029] In another possible implementation, the method further includes: in response to the user's sixth input on the second interface, sending a template save request to the serverless system to save the first parameter and the first cost value as a third resource configuration template. This allows the user to easily access this resource configuration template at any time, simplifying user operations and reducing the computational process of the serverless system.

[0030] In a third aspect, a resource configuration method is provided. The method is executed in a serverless system and includes: in response to a template request from a terminal, sending a first resource configuration template to the terminal, wherein the first resource configuration template includes second parameters and a second cost value, the second parameters including a set of second attribute parameters of one or more of scaling and / or scheduling a workload to be deployed, the computing resource type of the workload, or the workload's tolerance to interruptions, and the second cost value being a cost value of resources estimated based on the second parameters for deploying the workload; and in response to a configuration request from the terminal, configuring resources for deploying the workload according to the second parameters.

[0031] For example, the first resource configuration template can be a predefined template or a user-defined template, and this invention does not limit it.

[0032] For example, a configuration request may include a second parameter or an identifier for a first resource configuration template. Using this identifier, the serverless system can quickly identify a template containing information about the resources on which the user wants to deploy workloads.

[0033] In the solution provided in the third aspect above, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates) to users via a terminal, allowing users to directly select a template from the existing options. The serverless system can then present the user with corresponding parameters and cost values ​​based on the selected template for reference. This allows users to determine a workload deployment plan that meets their needs and preferences, simplifying user operations and reducing the computational process of the serverless system.

[0034] In a fourth aspect, a resource configuration method is provided. The method is executed in a terminal and includes: sending a template request to a serverless system; receiving from the serverless system a first resource configuration template including a second parameter and a second cost value, wherein the second parameter includes a set of second attribute parameters comprising one or more of scaling and / or scheduling a workload to be deployed, a computing resource type of the workload, or the workload's tolerance to interruptions, and the second cost value is a cost value estimated based on the second parameter for deploying resources for the workload; and, in response to user input, sending a configuration request to the serverless system, wherein the configuration request is for requesting the deployment of resources for the workload according to the second parameter.

[0035] In the solution provided in the fifth aspect above, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates) to users via a terminal, allowing users to directly select a template from the existing options. The serverless system can then present the user with corresponding parameters and cost values ​​based on the selected template for reference. This allows users to determine a workload deployment plan that meets their needs and preferences, simplifying user operations and reducing the computational process of the serverless system.

[0036] In a fifth aspect, a serverless system is provided. The serverless system includes: a transceiver for transmitting and receiving communication signals; a memory for storing computer program instructions; and a processor for executing the computer program instructions, such that the serverless system implements the method provided in any of the possible implementations of the first or third aspect.

[0037] In a sixth aspect, a resource allocation apparatus is provided. The apparatus includes: a transceiver for transmitting and receiving communication signals; a memory for storing computer program instructions; and a processor for executing the computer program instructions to cause the serverless system to implement the method provided in any of the possible implementations of the first or third aspect described above.

[0038] In a seventh aspect, a terminal device is provided. The terminal includes: a transceiver for transmitting and receiving communication signals; a memory for storing computer program instructions; and a processor for executing the computer program instructions to cause the serverless system to implement the method provided in any of the possible implementations of the second or fourth aspect described above.

[0039] In an eighth aspect, a serverless system is provided. The serverless system includes: a configuration manager, an application programming interface (API) server, a scheduler, and a scaler. The API server is configured to: receive a first request including a first parameter, wherein the first parameter includes a first set of attribute parameters comprising one or more of the following: the speed at which a workload to be deployed is scaled and / or scheduled; the type of computing resources of the workload; or the workload's tolerance to interruption; estimate a first cost value based on the first parameter; and output the first cost value. The configuration manager is configured to instruct the scheduler, upon receiving first feedback of the first cost value, to deploy resources for the workload based on the first parameter. The scheduler is configured to deploy resources for the workload according to the instructions from the configuration manager. The scaler is configured to prepare the resources required by the scheduler to deploy the workload.

[0040] In the solution provided in the eighth aspect above, the serverless system can offer configuration options, allowing users to define one or more attribute parameters such as the speed of scaling and / or scheduling workloads, the type of computing resources used for the workload, or the workload's tolerance to interruptions. This supports users in making detailed and flexible configurations or adjustments to resources based on the actual workload and their own preferences, achieving reasonable resource allocation. Furthermore, the serverless system can estimate and display the cost of the required resources based on the user's attribute configuration, intuitively informing the user of the costs they will subsequently incur. This allows users to decide whether to accept the workload deployment plan after receiving payment information. In this way, the serverless system can develop a scientific, reliable workload deployment plan that meets the user's needs, taking into account their specific circumstances.

[0041] In one possible implementation, the API server is further configured to: before receiving the first request, in response to a template request, output a resource configuration template including a second parameter and a second cost value, wherein the first parameter is a parameter obtained by adjusting the second parameter, and the second parameter includes a set of second attribute parameters comprising one or more of the following: the speed at which the workload is scaled and / or scheduled, the type of computing resources used by the workload, or the workload's tolerance to interruptions. In this way, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates). The workload can be deployed based on a workload profile corresponding to a template selected directly by the user from existing templates, simplifying user operations and reducing computation on the serverless system.

[0042] In another possible implementation, the API server is further configured to output first cost recommendation information, which indicates the impact of the difference between the first parameter and the second parameter on the cost of resources required to deploy the workload. This allows users to easily and intuitively understand the cost impact of different parameter sets, enabling them to tailor a more scientific, reliable, and user-friendly workload deployment plan based on their specific circumstances, thereby improving the user experience.

[0043] In another possible implementation, the attribute parameter for scaling and / or scheduling the workload includes the proportion of the workload to be deployed on at least one type of node among hot pool nodes, cold pool nodes, or instant pool nodes, wherein scaling and / or scheduling the hot pool nodes is faster than scaling and / or scheduling the cold pool nodes, and scaling and / or scheduling the cold pool nodes is faster than scaling and / or scheduling the instant pool nodes. In this way, the serverless system can support users in selecting different types of node resources based on the speed requirements of scaling and / or scheduling the workload, thereby developing a more scientific, reliable, and user-responsive workload deployment plan.

[0044] In another possible implementation, the attribute parameters of the computing resource type of the workload include at least one of input / output (I / O) intensive resources, central processing unit (CPU) intensive resources, or memory intensive resources. In this way, the serverless system can allow users to select different types of computing resources based on the specific needs of the workload in one or more aspects such as I / O performance, CPU performance, or memory performance, thereby developing a more scientific, reliable, and user-responsive workload deployment plan.

[0045] In another possible implementation, the workload's tolerance to interruption includes the following attribute parameters: the proportion of the workload to be deployed on on-demand instances, and / or the proportion of the workload to be deployed on spot instances. This allows serverless systems to support users in selecting instances with different levels of reliability based on their varying tolerance for interruption, thereby enabling more scientific, reliable, and user-responsive workload deployment solutions.

[0046] In another possible implementation, the API server is further configured to: estimate a third cost value in response to second feedback on the first cost value, wherein the second feedback includes a third parameter, the third parameter comprising a set of third attribute parameters including one or more of the scaling and / or scheduling speed of the workload, the computing resource type of the workload, or the workload's tolerance to interruptions; and output the third cost value. The configuration manager, the scheduler, and the scaler are further configured to configure resources for deploying the workload according to the third parameters after receiving third feedback on the third cost value. In this way, if the user is not satisfied with the first cost value, the serverless system allows the user to adjust attributes to obtain a workload deployment scheme that meets the user's cost expectations, thereby obtaining an alternative workload deployment scheme that satisfies the user's needs and preferences.

[0047] In another possible implementation, the API server is further configured to output second cost recommendation information, which indicates the impact of the difference between the first parameter and the third parameter on the cost of resources required to deploy the workload. This allows users to easily and intuitively understand the cost impact of different parameter sets, enabling them to tailor a more scientific, reliable, and user-friendly workload deployment plan based on their specific circumstances, thereby improving the user experience.

[0048] In a ninth aspect, a serverless system is provided. The serverless system includes: a configuration manager, an API server, a scheduler, and a scaler. The API server is configured to: in response to a template request from a terminal, send a first resource configuration template to the terminal, wherein the first resource configuration template includes a second parameter and a second cost value, the second parameter including a set of second attribute parameters of one or more of scaling and / or scheduling a workload to be deployed, the computing resource type of the workload, or the workload's tolerance to interruptions, and the second cost value being a cost value estimated based on the second parameter for deploying resources for the workload. The configuration manager is configured to instruct the scheduler, in response to a configuration request from the terminal, to deploy resources for the workload according to the second parameter. The scheduler is configured to deploy resources for the workload according to the instructions from the configuration manager. The scaler is configured to prepare the resources required by the scheduler to deploy the workload.

[0049] In the solution provided in the ninth aspect above, the serverless system can provide templates (e.g., predefined templates and / or user-defined templates) to users via a terminal, allowing users to directly select a template from the existing options. The serverless system can then present the user with corresponding parameters and cost values ​​based on the selected template for reference. This allows users to determine a workload deployment plan that meets their needs and preferences, simplifying user operations and reducing the computational process of the serverless system.

[0050] In a tenth aspect, a resource allocation system is provided. The system includes the serverless system provided in the fifth, eighth, or ninth aspects above, and the terminal device provided in the seventh aspect above.

[0051] In the eleventh aspect, a resource allocation system is provided. The system includes the resource allocation device provided in the sixth aspect and the terminal equipment provided in the seventh aspect.

[0052] In a twelfth aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores computer program instructions that, when executed by processing circuitry in a computer, cause the computer to implement the methods provided in any one of the first, second, third, and fourth aspects.

[0053] In a thirteenth aspect, a computer program product is provided. The computer program product includes instructions that, when executed by a computer, cause the computer to implement the method provided by any of the possible implementations of the first, second, third, and fourth aspects described above.

[0054] In a fourteenth aspect, a chip system is provided. The chip system includes processing circuitry and a storage medium, wherein the storage medium stores computer program instructions that, when executed by the processing circuitry, cause the chip system to implement the methods provided in any of the possible implementations of the first, second, third, and fourth aspects described above. The chip system includes one or more chips, or may include one or more chips and other discrete devices. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of resource management on several computing platforms; Figure 2 This is a schematic diagram of the resource configuration interface; Figure 3 This is a first schematic diagram of the resource allocation system provided in an embodiment of the present invention; Figure 4 This is a first schematic diagram of the resource configuration interface provided in an embodiment of the present invention; Figure 5 This is a first flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 6 This is a second schematic diagram of the resource configuration interface provided in an embodiment of the present invention; Figure 7 This is a third schematic diagram of the resource configuration interface provided in an embodiment of the present invention; Figure 8A This is a fourth schematic diagram of the resource configuration interface provided in this embodiment of the invention; Figure 8B This is a first schematic diagram of node deployment provided in an embodiment of the present invention; Figure 8C This is a second schematic diagram of node deployment provided in an embodiment of the present invention; Figure 8D This is a third schematic diagram of node deployment provided in an embodiment of the present invention; Figure 9 This is a second flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 10 This is a first schematic diagram of the attribute adjustment interface provided in an embodiment of the present invention; Figure 11 This is a second schematic diagram of the attribute adjustment interface provided in an embodiment of the present invention; Figure 12 This is a first schematic diagram of the cost value display interface provided in an embodiment of the present invention; Figure 13 This is a second schematic diagram of the cost value display interface provided in an embodiment of the present invention; Figure 14This is a first schematic diagram of the cost value and cost suggestion information display interface provided in the embodiments of the present invention; Figure 15 This is the third flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 16 This is a first schematic diagram of the template saving trigger interface provided in an embodiment of the present invention; Figure 17 This is the fourth flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 18 This is a second schematic diagram of the template saving trigger interface provided in the embodiment of the present invention; Figure 19 This is a second schematic diagram of the resource allocation system provided in an embodiment of the present invention; Figure 20 This is the fifth flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 21 This is a schematic diagram of a template selection scenario provided by an embodiment of the present invention; Figure 22 This is a schematic diagram of another template selection scenario provided by an embodiment of the present invention; Figure 23 This is a third schematic diagram of the cost value display interface provided in an embodiment of the present invention; Figure 24 This is a fourth schematic diagram of the cost value display interface provided in an embodiment of the present invention; Figure 25 This is a second schematic diagram of the cost value and cost suggestion information display interface provided in the embodiments of the present invention; Figure 26 This is the sixth flowchart of the resource allocation method provided in the embodiments of the present invention; Figure 27 This is the fifth schematic diagram of the cost value display interface provided in the embodiment of the present invention; Figure 28 This is a block diagram of the structure of the serverless resource configuration system or device provided in the embodiments of the present invention; Figure 29 This is a block diagram of the structure of the terminal device provided in the embodiments of the present invention. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the embodiments of the present invention, unless otherwise stated, the symbol " / " represents "or," for example, "A / B" can represent A or B; "and / or" in this document only indicates the relationship between related objects, indicating that there can be three relationships, for example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. Furthermore, in the description of the embodiments of the present invention, "one / more" refers to two or more.

[0057] In the following text, the terms "first," "second," etc., are used only to distinguish different descriptive objects and do not limit the position, order, priority, number, or content of the descriptive objects. For example, if the descriptive object is a "field," the ordinal numbers before "field" in "first field" and "second field" do not restrict the position or order of the "fields," nor do "first" and "second" restrict whether the modified "fields" are in the same message, and the order of "first field" and "second field" is also unrestricted. Similarly, if the descriptive object is a "level," the ordinal numbers before "level" in "first level" and "second level" do not restrict the priority between the "levels." For example, the number of descriptive objects is not limited by their ordinal numbers; there can be one or more descriptive objects. Taking "first device" as an example, the number of "devices" can be one or more. Furthermore, objects modified by different prefixes can have the same or different types / content. For example, if the object being described is "device," then "first device" and "second device" can be devices of the same type or devices of different types; similarly, if the object being described is "information," then "first information" and "second information" can be information with the same content or information with different content. In short, the use of ordinal numbers and other prefixes in the embodiments of this invention does not limit the object being described. The description of the object should refer to the contextual description in the claims or embodiments, and should not constitute unnecessary limitations due to the use of these prefixes.

[0058] Furthermore, in this embodiment of the invention, "connection" can be a direct connection or an indirect connection, and can also refer to an electrical connection or a communication connection. For example, two electrical components A and B are connected to each other. This can mean that A and B are directly connected, or that A and B are indirectly connected through other electrical components or connection media, or that A and B are indirectly connected through other communication devices or communication media, as long as A and B can communicate with each other.

[0059] Currently, such as Figure 1As shown, a user's application / workload (hereinafter referred to as "workload") can be deployed on a server computing platform (server computer) or a serverless computing platform (serverless computer). For example, server computing platforms include Elastic Kubernetes Service (EKS), Azure Kubernetes Service (AKS), Kubernetes Engine (KE), Cloud Container Engine (CCE), EKS with autoscaling, Azure Kubernetes Service with autoscaling, Google Kubernetes Engine (GKE) with cluster scaling, and CCE with scaling, etc.; while serverless computing platforms include Web service Fargate, Azure Container App, Cloud Run, Elastic Container Instance (ECI), Cloud Container Instance (CCI), CCI-Elastic Cloud Server (CCI-ECS), etc.

[0060] like Figure 1 As shown, server computing platforms can be jointly managed and maintained by equipment manufacturers and equipment suppliers, i.e., "customer-vendor management", or managed by equipment suppliers, i.e., "customer management"; while serverless computing platforms are usually provided by equipment manufacturers, i.e., "fully vendor-managed", meaning that the equipment manufacturers are responsible for management and maintenance.

[0061] Here, Figure 1 The CCI-ECS, EKS, AKS, KE, and CCE shown are high-configuration platforms; Figure 1 The web service Fargate, EKS with cluster autoscaler, Azure Kubernetes Service with cluster autoscaler, GKE with cluster scaling, and CCE with scaling are limited configuration platforms. Figure 1 The Azure Container App, Cloud Run, ECI, and CCI shown are non-configurable platforms.

[0062] In server computing platforms, users first purchase infrastructure resources (such as VMs or Kubernetes clusters) and then deploy their workloads on those resources. Costs are incurred throughout this process, regardless of whether the user actually uses the purchased resources. In contrast, serverless computing platforms allow users to pay for the resources they use only after deploying their workloads, thus significantly reducing user costs.

[0063] The embodiments of this invention mainly relate to the following situation: a serverless computing platform provides users with resources for deploying workloads. As mentioned above, a serverless computing platform can allow users to deploy applications / workloads (hereinafter referred to as "workloads"), for example, the serverless computing platform can provide users with a channel to purchase infrastructure resources (e.g., VM or Kubernetes cluster resources) for deploying their workloads.

[0064] like Figure 2 As shown, existing serverless computing platforms can provide an interface 200 for resource configuration. Based on the amount of hardware resources required by the workload, such as the number of CPUs, memory size, number of deployed replicas, and number of spot instances, users can flexibly configure or adjust resources on the interface 200 as needed. For example, they can configure or adjust the number of tasks or spots, usage duration (e.g., daily), average duration, allocated storage resources, and storage resources temporarily allocated to ECS instances, etc. Here, a replica is a copy of the workload / application to be deployed; users typically deploy multiple copies / replicas to distribute the workload among them. To help users understand the cost of the resources required to deploy their workloads, the interface 200 can help users understand the relevant attributes of different types of resources by displaying a compute item drop-down menu and corresponding documentation. Based on this, users can estimate the cost required to deploy their workloads on the serverless computing platform.

[0065] The workloads described in these embodiments may include, but are not limited to, containerized application workloads, non-containerized application workloads, microservice workloads, and batch application workloads. These embodiments do not specifically limit the workload; the workload may depend on the specific use case and business authorization requirements.

[0066] As is well known, different workloads require different specific resources. For example, some workloads require CPU-intensive resources, some require memory-intensive resources, and some require low-interrupt-tolerance resources. However, the resource configuration interfaces provided by existing server computing platforms (hereinafter referred to as "resource configuration interfaces") do not offer users more detailed options. Furthermore, while serverless computing platforms provide drop-down menus and documentation to help users understand the attributes of different resource types, those lacking a basic understanding of infrastructure resources (e.g., resource attributes) will find it difficult to grasp the cost impact of different resource configurations. Additionally, the drop-down menus provided by server computing platforms require users to navigate through various dropdowns and selections to view detailed attribute information, some of which may be redundant and difficult for users to navigate. Therefore, the operation is extremely inconvenient for users, and they cannot accurately obtain the information they want. Consequently, existing serverless computing platforms cannot provide users with an intuitive and detailed cost breakdown of resource configuration options.

[0067] To provide an intuitive and detailed resource configuration scheme so that users can configure or adjust detailed and flexible resources according to their actual workload resource requirements and preferences, and intuitively understand the impact of different resource parameters on costs, this invention provides a method for configuring serverless resources. This method can be implemented based on the workload-aware configuration manager (hereinafter referred to as the "configuration manager") of the serverless computing platform (also known as the "serverless system").

[0068] In one possible example Figure 3 This is a schematic diagram of the resource allocation system provided in an embodiment of the present invention. For example... Figure 3 As shown, a resource configuration system may include a configuration manager (e.g., a load-aware configuration manager), a user interface (UI) display device, an application programming interface (API) server, a scheduler, and a scaler.

[0069] The configuration manager is primarily used to configure workload attributes or resource scaling attributes based on user actions.

[0070] Here, workload attribute configuration may include, but is not limited to, resource preference configuration, such as node preference configuration, resource type preference configuration, and outage tolerance preference configuration.

[0071] Nodes can include, but are not limited to, instant pool nodes, cold pool nodes, and hot pool nodes. These are concepts within a "hot-warm" architecture. Hot pool nodes are the fastest, followed by cold pool nodes, with instant pool nodes being the slowest. Hot pool nodes are typically used to store data that is most important or has high priority to users. Instant pool nodes and cold pool nodes are typically used to store data that is not particularly important or has low priority to users. In other words, scaling and / or scheduling a portion of workload to be deployed on hot pool nodes is faster than scaling and / or scheduling a portion of workload to be deployed on cold pool nodes, and scaling and / or scheduling a portion of workload to be deployed on cold pool nodes is faster than scaling and / or scheduling a portion of workload to be deployed on instant pool nodes.

[0072] Resource types may include, but are not limited to, computing resource types and storage resource types. Computing resource types may include, but are not limited to, one or more of the following: general-purpose computing resources, general-purpose computing enhanced resources, memory-optimized resources, large memory resources, high-performance computing resources, disk-intensive resources, ultra-high input / output (I / O) resources, or graphics processing unit (GPU) accelerated resources.

[0073] In some embodiments, the resource type can be a specific resource type under the central processing unit (CPU) architecture. The CPU architecture can include, but is not limited to, the x86 CPU architecture, the Kunpeng architecture, etc., and the embodiments of the present invention do not specifically limit it.

[0074] For example, an outage tolerance preference configuration could be to deploy workloads on spot instances or on-demand instances. Here, spot instances are VM instances that utilize the cloud's available capacity. These instances are typically sold at a discount. However, when workload demand increases, these spot instances are reclaimed, and their reliability decreases; on-demand instances, on the other hand, are more reliable than spot instances, and therefore more expensive. Given this, if all workloads cannot tolerate outages, then 100% of the workload will be deployed on on-demand instances; if all workloads can tolerate outages, then the workloads will be deployed on spot instances; if the workloads can tolerate partial outages, then a1% of the workload will be deployed on on-demand instances, and a2% of the workload will be deployed on spot instances, where a1% + a2% = 100%. Here, on-demand instances are more expensive than spot instances.

[0075] For example, resource scaling attribute configurations may include, but are not limited to, scaling speed and / or scheduling speed. For example, scaling speed and / or scheduling speed configurations can be achieved through node scaling configurations such as scalable instant pool nodes, cold pool nodes, or hot pool nodes.

[0076] UI display devices are human-computer interfaces used to provide users with workload attribute configuration or resource scaling attribute configuration, such as resource configuration interfaces.

[0077] In some embodiments, the UI display device can also be used to display to the user the cost of resources corresponding to the operations configured with workload attributes or resource scaling attributes through a resource configuration interface.

[0078] In some embodiments, the UI display device can also be used to display adjustments made by the user to workload attributes or resource scaling attributes based on cost.

[0079] For example, the UI display device can be located on the user side. For instance, the user can connect to the configuration manager via an API server through a terminal device such as a personal computer (PC) or tablet. The terminal device then receives interface data from the resource configuration interface from the configuration manager and displays the corresponding UI. The UI display device can also provide feedback to the configuration manager based on the operations configured in the workload attribute configuration or resource scaling attribute configuration on the resource configuration interface.

[0080] For example, the UI display device can be located on the serverless computing platform side. The configuration manager can display the resource configuration interface through the UI display device. The UI display device can also provide feedback to the configuration manager based on the workload attribute configuration or resource scaling attribute configuration operations on the resource configuration interface.

[0081] The API server is used to calculate the corresponding costs based on the operations configured in the workload attribute settings or resource scaling attribute settings on the resource configuration interface, and to display the costs through the resource configuration interface.

[0082] Based on the user's workload attribute configuration or resource scaling attribute configuration and following the instructions of the configuration manager, the scheduler and scaler are used to provide resources for the configuration. For example, the scheduler may instruct the scaler to warm up and / or start the required resources according to the user's workload attribute configuration or resource scaling attribute configuration, and to deploy the workload using the resources warmed up and / or started by the scaler.

[0083] The method for configuring serverless resources provided by the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0084] In some embodiments, when a user wants to deploy a workload on a serverless system (e.g., Huawei Cloud CCI serverless computing platform), the serverless system can provide a resource configuration interface for users to configure standard options for workload metadata, such as... Figure 4 The workload name, namespace, description, number of instances, instance specifications, and size (e.g., CPU size, memory size) are shown.

[0085] Figure 5 This is a flowchart of a resource configuration method provided in an embodiment of the present invention.

[0086] like Figure 5 As shown, taking the method for configuring serverless resources applied to a serverless system as an example, the serverless system includes... Figure 3 The configuration manager, API server, scheduler, and scaler shown in the embodiments of the present invention can include S501 to S504 as the resource configuration method provided.

[0087] In S501, the serverless system receives the first request.

[0088] The first request includes a first set of parameters, which includes one or more of the following first attribute parameters: the speed at which the workload to be deployed is scaled and / or scheduled; the compute resource type of the workload; and the workload's tolerance to interruption. The workload's tolerance to interruption may be expressed as the ratio of a portion of the workload to be deployed on an on-demand instance to the total workload, and / or the ratio of a portion of the workload to be deployed on a spot instance to the total workload.

[0089] In some embodiments, the first request is sent by the UI display device to the serverless system. For example, the UI display device may send a set of parameters (e.g., a first parameter) corresponding to one or more attributes to the serverless system via the first request, based on a first input from a user on a first interface. For example, the first interface may be a resource configuration interface; the first input may be an operation to configure or adjust one or more attributes (e.g., the speed of scaling and / or scheduling workloads, the type of computing resources for the workload, or the workload's tolerance to interruptions).

[0090] For example, the UI display device can be a terminal-side device, and a wired or wireless communication connection can be established between the UI display device and the serverless system. Based on configuration or adjustment operations performed by the user on one or more attributes (e.g., the speed of scaling and / or scheduling workloads, the type of computing resources for the workload, or the workload's tolerance to interruptions) on the resource configuration interface, the UI display device can send a set of parameters (e.g., a first parameter) corresponding to one or more attributes to the serverless system via the wired or wireless communication connection between the UI display device and the serverless system through a first request.

[0091] For example, the UI display device can be a device on the serverless system side. The UI display device can be used as a display device in the serverless system to connect to the serverless system via a wired or wireless communication connection. Based on the configuration or adjustment operations made by the user on one or more attributes (e.g., the speed of scaling and / or scheduling workloads, the computing resource type of the workload, or the workload's tolerance to interruptions) on the resource configuration interface, the UI display device can send a set of parameters (e.g., a first parameter) corresponding to one or more attributes to the serverless system via a wired or wireless communication connection between the UI display device and the serverless system through a first request.

[0092] For example, the resource configuration interface can be like this Figure 6 As shown. Reference Figure 6 The resource configuration interface can include scaling / scheduling preferences, compute resource type options, and guaranteed replication options. Scaling / scheduling preferences and compute resource type options are used to configure workload attributes and / or resource scaling attributes. For example, scaling / scheduling preferences can be used to configure the speed of scaling and / or scheduling workloads, compute resource type options can be used to configure the compute resource type of the workload, and guaranteed replication options can be used to configure the workload's tolerance to interruptions.

[0093] The scaling / scheduling preference options may include, but are not limited to, one or more of instant pool nodes, hot pool nodes, or cold pool nodes; the compute resource type options may include, but are not limited to, one or more of I / O intensive resources, CPU intensive resources, or memory intensive resources. I / O intensive types may refer to ultra-high I / O resources; CPU intensive types may refer to general-purpose compute-enhanced resources, general-purpose compute resources, GPU-accelerated resources, high-performance compute resources, etc.; memory intensive types may refer to memory-optimized resources, large memory resources, disk-intensive resources, etc. The guaranteed replication option may refer to the proportion of workload to be deployed on on-demand instances and / or the proportion of workload to be deployed on spot instances. Guaranteed replication options may include, but are not limited to, one or more of 100%, 75%, 70%, 50%, 30%, 25%, or 0%.

[0094] In some embodiments, the resource configuration interface may also include extensions, such as... Figure 6 The future parameters are shown here. Figure 6 Only future parameter 1 and future parameter 2 are shown; this embodiment of the invention does not limit the number of future parameters. Therefore, if the serverless system supports configuring other related attributes in the future, the corresponding attributes can be directly configured based on the extended items.

[0095] In some embodiments, such as Figure 6 As shown, in addition to multiple configuration items including scaling / scheduling preferences, compute resource type items, and guaranteed replication items, the resource configuration interface can also include a first view. The first view can display the first parameters geometrically according to the user's configuration or adjustment operations on one or more attributes, thereby enhancing parameter visualization and improving the user experience.

[0096] Here, for example, the first view (and the other views below) can be Figure 6 The diagram shows a pentagon with points and lines. Of course, this invention does not limit the specific graphic form of the view, as long as the view can display the corresponding parameters in geometric form.

[0097] In some embodiments, such as Figure 6 As shown, the resource configuration interface can also display reference lines at the location of the view (e.g., the first view). Figure 6The reference lines (shown as multiple regular pentagons) are used to help users understand the adjustable range of points for various attributes. They also facilitate user experience by allowing users to adjust one or more attributes (e.g., scaling / scheduling preferences, compute resource types, and guaranteed replication) by dragging points of one or more configuration items in the view. Similarly, other interfaces in this embodiment can also display reference lines corresponding to views for users to adjust attributes. For example, a user can first select the corresponding view, for example, hover the mouse or touch point over the lines of the view, and then drag points in the view to adjust it.

[0098] In some embodiments, after a user completes configuration or adjustment operations on one or more attributes, the UI display device can send a set of parameters (i.e., first parameters) corresponding to one or more attributes to the serverless system via a first request. For example, the UI display device can automatically trigger the operation of sending the first parameters to the serverless system after the user completes configuration or adjustment operations on one or more attributes. Alternatively, the UI display device can trigger the operation of sending the first parameters to the serverless system after the user completes configuration or adjustment operations on one or more attributes and clicks the virtual button "FinishConfiguration" or "OK" on the resource configuration interface. This embodiment of the invention does not limit the specific time and triggering method of the UI display device sending the first parameters to the serverless system.

[0099] It should be noted that, Figure 6 The interface shown is merely an example. This embodiment of the invention does not limit the specific number and content of the pre-configured types of user-selectable attribute parameters (described below), templates (also described below), and configuration items. For example, Figure 6 The interface shown may also include a side scrollbar; as the user scrolls down, a scrolling effect appears on the screen, displaying other configuration items. For example, Figure 6 The interface shown may also include a preference ratio for computing resources, for example, 100% for instant pool nodes and 100% for hot pool nodes, or 50% for cold pool nodes and 50% for instant pool nodes.

[0100] In S502, the serverless system estimates a first cost value based on a first parameter.

[0101] In some embodiments, the serverless system can estimate a first cost value based on a first parameter through an API server. The first cost value may be the unit time cost of the instance required by the first parameter, for example, $12 per instance per hour. This embodiment of the invention does not limit the specific estimation methods and processes for the first cost value and other cost values ​​below; for specific estimation methods and processes, please refer to the prior art.

[0102] In some embodiments, the serverless system can also preheat and / or start resources in a resource pool according to a first parameter via a scaler, so that workloads can be subsequently deployed on the corresponding resources in the resource pool. For example, suppose the first input is an operation to determine the speed of scaling workloads, such as a user choosing to deploy 50% of the workload on cold pool nodes and another 50% on instant pool nodes. In this case, the scaler can preheat resources already running in the resource pool according to the first parameter, for example, preheating the required number of cold pool nodes, and once these nodes are ready, the scheduler will deploy the corresponding proportion of workloads on these nodes; the scaler can also create the required number of instant pool nodes according to the first parameter, for example, by starting some resources in the resource pool, and once these nodes are running and available, the scheduler will deploy the corresponding proportion of workloads on these nodes.

[0103] In S503, the serverless system outputs the first cost value.

[0104] In some embodiments, the serverless system outputting a first cost value may include the serverless system displaying the first cost value through a UI display device. For example, an API server in the serverless system may send an estimated first cost value to the UI display device so that the UI display device displays the first cost value to a user through an interface (e.g., a second interface).

[0105] For example, UI display devices can be used Figure 7 The interface shown (i.e., via the second interface) displays a first cost value to the user, such as the required cost. For example, the cost is $12 per copy per hour.

[0106] In some embodiments, the UI display device can also display cost details corresponding to the cost values ​​through the interface. Figure 7 Taking the second interface, which displays the first cost value, as an example, if the user clicks the virtual button "View details" on the second interface, the cost details corresponding to the first cost value will be displayed on the second interface. For example... Figure 7As shown, taking the first parameter including scaling / scheduling preference as instant pool node, computing resource type as I / O intensive resource, and guaranteed replication as 50% as an example, the cost details corresponding to the first cost value include: the cost value corresponding to the instant pool node (e.g., $4 per hour), the cost value corresponding to the I / O intensive resource (e.g., $3 per hour), and the cost value corresponding to 50% guaranteed replication (e.g., $5 per hour). Here, other cost values ​​in the embodiments of the present invention can be shown in the form of detailed cost values ​​using similar cost details.

[0107] In S504, after receiving the first feedback on the first cost value, the serverless system configures resources for deploying workloads according to the first parameter.

[0108] For example, the first feedback on the first cost value can determine the deployment of workloads based on the resources used for the first parameter corresponding to the first cost value.

[0109] It's important to understand that different workloads require different specific resources. For example, some workloads require CPU-intensive resources, some require memory-intensive resources, and some can tolerate interruptions. In addition, different users have different priorities. For example, some users are more concerned with low cost, while others are more concerned with the execution speed or reliability of the workload, and so on.

[0110] For example, if a user is satisfied with the first cost value—that is, if the user can afford the cost corresponding to the first cost value and decides to adopt the workload deployment plan—the user can trigger the UI display device to send the first feedback on the first cost value to the serverless system.

[0111] In some embodiments, upon receiving a second input from a user on a second interface, the UI display device may send a first feedback to the serverless system regarding a first cost value, wherein the first feedback is used to instruct the configuration of resources for deploying workloads according to a first parameter.

[0112] For example, when the UI display device receives a user click Figure 8A After clicking the "Confirm Deployment" virtual button on the interface shown (i.e., the second interface), the UI display device can send initial feedback on the first cost value to the configuration manager in the serverless system. Upon receiving the initial feedback on the first cost value, the configuration manager in the serverless system can instruct the scheduler to configure the resources used to deploy the workload according to the first parameter.

[0113] For example, Figure 8B This is a schematic diagram illustrating a node deployment according to an embodiment of the present invention. Figure 8BAs shown, the scheduler can deploy workloads on one or more of the instant pool nodes, cold pool nodes, or hot pool nodes based on the scaling / scheduling preferences in the first parameter.

[0114] For example, if the scaling / scheduling preference is for an instant pool node (meaning the workload has no speed requirements), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on instant pool nodes, or deploy the majority of workloads on instant pool nodes, to minimize costs. Similarly, if the scaling / scheduling preference is for a hot pool node (meaning the workload requires high speed), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on hot pool nodes to minimize costs; alternatively, the configuration manager can instruct the scheduler to deploy the majority of workloads on hot pool nodes to minimize costs while ensuring workload speed. Finally, if the scaling / scheduling preference is for a cold pool node (meaning the workload requires medium speed), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on cold pool nodes, or deploy the majority of workloads on cold pool nodes, to minimize costs while ensuring workload uptime.

[0115] For example, Figure 8C This is a schematic diagram illustrating another node deployment provided by an embodiment of the present invention. For example... Figure 8C As shown, the scheduler can deploy workloads on one or more of the following types of computing resources, such as general computing enhanced resources, general computing resources, GPU accelerated resources, high-performance computing resources, memory-optimized resources, large memory resources, and disk-intensive resources, based on the computing resource type in the first parameter.

[0116] For example, if the compute resource type is set to I / O intensive (meaning the workload requires high I / O performance), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on nodes that meet ultra-high I / O standards to satisfy the workload's I / O performance requirements. Similarly, if the compute resource type is set to CPU intensive (meaning the workload requires high CPU computing power), the configuration manager can instruct the scheduler to deploy all workloads on one or more of general-purpose compute-enhanced resources, high-performance compute resources, or GPU-accelerated resources to satisfy the workload's CPU computing power requirements. For instance, all workloads could be deployed on one or more of C1, C2, S1, S2, S3, or Sn3 resources. For example, if the user selects artificial intelligence (AI) inference compute resource type, the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on GPU-accelerated resources (e.g., GPU-accelerated nodes). For example, if the compute resource type is memory-intensive (meaning the workload requires high storage performance), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on one or more of memory-optimized resources, large memory resources, or disk-intensive resources.

[0117] For example, Figure 8D This is a schematic diagram illustrating another node deployment method provided in an embodiment of the present invention. For example... Figure 8D As shown, the scheduler can deploy workloads on on-demand instances and / or spot instances based on the guaranteed replica in the first parameter.

[0118] To ensure 100% replication (meaning the workload cannot tolerate interruptions), the configuration manager in a serverless system can instruct the scheduler to deploy all workloads on replicas scheduled via on-demand instances, minimizing workload downtime. Conversely, to ensure 0% replication (meaning the workload has a high tolerance for instance downtime), the configuration manager can instruct the scheduler to deploy all workloads on replicas scheduled via spot instances, minimizing costs. And to ensure 50% replication (meaning the workload can tolerate 50% instance downtime), the configuration manager can instruct the scheduler to deploy 50% of the workload on replicas scheduled via on-demand instances and the remaining 50% on replicas scheduled via spot instances, minimizing costs while still meeting the workload's tolerance for downtime. For example, if the replication factor is 70% (meaning the workload can tolerate 30% of the instances being interrupted), the configuration manager in a serverless system can instruct the scheduler to deploy 70% of the workload on replicas scheduled via on-demand instances and the remaining 30% on replicas scheduled via spot instances, in order to minimize costs while meeting the workload's tolerance for interruption.

[0119] In some embodiments of the present invention, if a user is not satisfied with the first cost value, that is, if the user believes that the cost corresponding to the first cost value is too high and decides to abandon the workload deployment plan, the user can trigger the UI display device to send a second feedback on the first cost value to the serverless system.

[0120] For example, if a user is not satisfied with the initial cost value, they can adjust the attribute parameters on the resource configuration interface displayed on the UI device to obtain an ideal workload deployment plan that meets their cost expectations. In this way, the UI device can send a second feedback on the initial cost value to the serverless system based on the user's adjustments to the attribute parameters. This second feedback includes a third parameter, which the serverless system can use to estimate a corresponding third cost value.

[0121] For example, Figure 9 This is a flowchart of another resource configuration method provided in an embodiment of the present invention.

[0122] like Figure 9 As shown, after S501 to S503, the resource configuration method provided in this embodiment of the invention may further include steps S901 to S904.

[0123] In S901, the serverless system receives a second feedback on the first cost value.

[0124] The second feedback includes a third parameter, which is a parameter obtained by adjusting the second parameter, and includes a set of third attribute parameters, such as the speed at which the workload is scaled and / or scheduled, the type of compute resources used by the workload, or the workload's tolerance to interruptions.

[0125] In some embodiments, the second feedback is sent from the UI display device to the serverless system. For example, the UI display device may send second feedback on the first cost value to the serverless system based on a fourth input from the user on the second interface. The fourth input on the second interface may be an adjustment operation made by the user on one or more items such as scaling / scheduling preferences, computing resource type items, and guaranteed replication items on the second interface. For example, Figure 10 This shows the user's action of adjusting the guaranteed replication factor to 70% on the resource configuration interface.

[0126] In some embodiments, a user can adjust one or more attributes by dragging points of one or more attributes in the first view, such as scaling / scheduling preferences, compute resource type, and guaranteed replication. Figure 11 This demonstrates how a user can adjust the guaranteed replication factor to 70% on the resource configuration interface by dragging a point on the guaranteed replication factor in the first view.

[0127] In some embodiments, such as Figure 10 As shown, when a user adjusts the guaranteed replication setting to 70% in the resource configuration interface, the view on the resource configuration interface will be updated synchronously, for example, from the first view to the third view. Additionally, as... Figure 11 As shown, when the user drags the point of the guaranteed replication item in the first view, the attribute parameters on the resource configuration interface will also be updated synchronously, for example, from the first parameter to the third parameter. The scaling / scheduling preference options may include, but are not limited to, one or more of instant pool nodes, hot pool nodes, or cold pool nodes; the compute resource type options may include, but are not limited to, one or more of I / O intensive resources, CPU intensive resources, or memory intensive resources; the guaranteed replication item options may include, but are not limited to, one or more of 100%, 75%, 70%, 50%, 30%, 25%, or 0%.

[0128] In some embodiments, the UI display device can send a second feedback to the serverless system after the user has adjusted one or more attributes. For example, the UI display device can automatically trigger the operation of sending the second feedback to the serverless system after the user has adjusted one or more attributes. Alternatively, the UI display device can trigger the operation of sending the second feedback to the serverless system after the user has adjusted one or more attributes and clicked the virtual button "Finish Configuration" or "OK" on the resource configuration interface. This embodiment of the invention does not limit the specific time and triggering method of the UI display device sending the second feedback to the serverless system.

[0129] In S902, the serverless system estimates the third cost value based on the third parameter.

[0130] In some embodiments, the serverless system can estimate the corresponding third cost value based on a third parameter through an API server. The third cost value can be the unit time cost of the replicas required by the third parameter, for example, $14 per replica per hour. This embodiment of the invention does not limit the specific estimation method and process of the third cost value; however, reference can be made to existing technologies regarding specific estimation methods and processes.

[0131] In S903, the serverless system outputs a third cost value.

[0132] In some embodiments, the serverless system outputting a third cost value may include displaying the third cost value via a UI display device. For example, the API server in the serverless system may send the estimated third cost value to the UI display device so that the UI display device can display the third cost value to the user through a fourth interface.

[0133] For example, UI display devices can be used Figure 12 The interface shown (i.e., the fourth interface) displays the third cost value to the user.

[0134] In some embodiments, the UI display device may also retain the previously received first cost value for the user to compare different cost values, allowing the user to select a preferred resource deployment scheme. For example... Figure 13 As shown, the UI display device can simultaneously display the cost value before attribute adjustment (the previous cost value) and the cost value after attribute adjustment, so that users can compare these cost values.

[0135] In some embodiments, such as Figure 13As shown, the fourth interface may also include a first view and a third view. The first view can display the first parameter in geometric shape according to the user's configuration or adjustment operations on one or more attributes, and the third view can display the third parameter in geometric shape according to the user's configuration or adjustment operations on one or more attributes, thereby enhancing the visualization of the parameters and improving the user experience.

[0136] In some embodiments, the serverless system may also output second cost suggestion information, which can then be displayed to the user by a UI display device. For example, Figure 14 This illustrates an example of a UI display device showing a second cost suggestion to a user. Figure 14 The second cost recommendation information shown includes suggestions for adopting a workload deployment scheme corresponding to the first cost value.

[0137] In some embodiments, the second cost recommendation information can also be used to indicate the impact of the difference between the first and third parameters on the cost of the resources required to deploy the workload; such as Figure 14 As shown, if the copy item is adjusted to 70%, the cost per unit time will increase by $2, allowing users to intuitively see the impact of different attribute combinations on costs.

[0138] In some embodiments, the second cost recommendation information may also include recommendations for adjusting attributes, such as adjusting an attribute to a certain parameter.

[0139] In S904, after receiving third feedback on the third cost value, the serverless system configures resources for deploying workloads based on the third parameter.

[0140] For example, third feedback on the third cost value can determine how to deploy workloads based on the resources used for the third parameter corresponding to the third cost value.

[0141] For example, if a user is satisfied with the third cost value—that is, if the user can afford the cost corresponding to the third cost value and decides to adopt the workload deployment plan—the user can trigger the UI display device to send a third feedback on the third cost value to the serverless system.

[0142] For example, when the UI display device receives a user click Figure 12 or Figure 13 After clicking the "Confirm Deployment" virtual button on the interface, the UI display device can send a third feedback on the third cost value to the configuration manager in the serverless system. Upon receiving this third feedback, the configuration manager in the serverless system can instruct the scheduler to configure the resources used to deploy the workload based on the third parameter.

[0143] In some embodiments of the present invention, if the user is dissatisfied with the third cost value—that is, if the user believes the cost corresponding to the third cost value is too high and decides to abandon the workload deployment plan—the user can adjust the attributes again to trigger the UI display device to send a fourth feedback on the third cost value to the serverless system. Then, the serverless system can estimate the cost value again based on the adjusted attribute parameters and output the cost value, which will not be elaborated further here. Alternatively, if the user compares the third cost value with the first cost value and ultimately decides to adopt the workload deployment plan corresponding to the first cost value, the user can adjust the attribute parameters back to the attribute parameter set corresponding to the previous first cost value, or directly adjust the attribute parameters back to the attribute parameter set corresponding to the previous first cost value by clicking the shortcut button on the resource configuration interface. The embodiments of the present invention do not limit the specific implementation methods and processes.

[0144] Understandably, according to Figure 5 or Figure 9 The resource configuration method shown allows the serverless system to provide configuration items that enable users to define scaling / scheduling preferences, compute resource types, or one or more attribute parameters for guaranteed replication. This allows users to configure their resource usage preferences according to their actual workload, achieving rational resource allocation. Furthermore, this method allows the serverless system to estimate and display the cost of required resources based on user attribute configurations, intuitively informing users of the costs they will subsequently incur. This allows users to decide whether to accept a workload deployment plan after receiving payment information. Moreover, this method allows the serverless system to adjust the displayed costs of different resources required for various workload deployment plans based on user attributes. Users can compare costs and other aspects by adjusting one or more attributes, allowing them to decide which workload deployment plan to adopt after receiving payment information. In this way, users can choose a workload deployment plan according to their own needs, thereby improving the user experience.

[0145] In some embodiments of the present invention, after a user completes the configuration or adjustment of one or more attributes and obtains the corresponding cost from the serverless system, the serverless system may also save the parameters and cost values ​​of the workload deployment scheme as a template according to a request to save the workload deployment scheme as a template for the user to select / use later.

[0146] For example, Figure 15 This is a flowchart of another resource configuration method provided in an embodiment of the present invention.

[0147] like Figure 15As shown, after executing S501 to S504, the resource configuration method provided in this embodiment of the invention may further include steps S1501 and S1502.

[0148] In S1501, the serverless system receives the second request.

[0149] The second request (e.g., a template save request) is used to request that the workload deployment scheme (including the first parameter and the first cost value) corresponding to the first cost value be saved as a template.

[0150] In some embodiments, the second request is sent from the UI display device to the serverless system. For example, the UI display device may send the second request to the serverless system based on a sixth input from the user on the second interface. The sixth input on the second interface may be an operation that saves a template including the first cost value on the resource configuration interface.

[0151] For example, serverless systems can display information through a UI display device. Figure 16 The interface shown includes a virtual "Save as template" button. It responds to user clicks. Figure 16 The virtual "Save as template" button on the interface shows that the UI display device can send a second request to the serverless system.

[0152] In S1502, the serverless system saves the first parameter and the first cost value as a first template according to the second request.

[0153] The first template saved by the serverless system can be directly invoked and then used to deploy workloads, as will be described in detail below.

[0154] For example, Figure 17 This is a flowchart of another resource configuration method provided in an embodiment of the present invention.

[0155] like Figure 17 As shown, after executing S501 to S503 and S901 to S904, the resource configuration method provided in this embodiment of the invention may further include steps S1701 and S1702.

[0156] In S1701, the serverless system receives a third request.

[0157] The third request is used to request that the workload deployment plan corresponding to the third cost value be saved as a template.

[0158] In some embodiments, the third request is sent by the UI display device to the serverless system. For example, the UI display device may send a third request to the serverless system based on a user's action of saving a template that includes a third cost value on the resource configuration interface.

[0159] For example, serverless systems can display information through a UI display device. Figure 18 The interface shown includes a virtual "Save as Template" button. It responds to user clicks. Figure 18 The virtual button "Save as template" on the interface shown allows the UI display device to send a third request to the serverless system.

[0160] In S1702, the serverless system saves the third parameter and the third cost value as a third template based on the third request.

[0161] The third template saved by the serverless system can be directly invoked and then used to deploy workloads.

[0162] In some embodiments of the present invention, such as Figure 19 As shown, the serverless system can also save and maintain one or more workload deployment templates. These templates can include, but are not limited to, predefined templates and user-defined templates. User-defined templates are those requested to be saved by the serverless system after the user has defined a set of attribute parameters; for example, templates saved by the serverless system based on the template operations displayed on the resource configuration saving interface.

[0163] In some embodiments of the present invention, the serverless system may provide one or more templates (including predefined templates and / or user-defined templates). Workloads can be deployed based on workload profiles corresponding to templates selected directly by the user from existing templates, simplifying user operations and reducing computation on the serverless system. Here, the cost value corresponding to the template can be recommended by the serverless system based on the type and / or content of the workload to be deployed.

[0164] For example, Figure 20 This is a flowchart of another resource configuration method provided in an embodiment of the present invention.

[0165] like Figure 20 As shown, the resource configuration method provided in this embodiment of the invention may include steps S2001 and S2002.

[0166] In S2001, the serverless system responds to template requests by outputting resource configuration templates.

[0167] The resource configuration template includes a second parameter and a second cost value. The first parameter is a parameter obtained by adjusting the second parameter. The second parameter includes a set of one or more of the following second attribute parameters: the speed at which the workload is scaled and / or scheduled, the type of compute resources used for the workload, or the workload's tolerance to interruptions.

[0168] In some embodiments, the template request is sent from the UI display device to the serverless system.

[0169] For example, a UI display device can send a template request to a serverless system based on a third input from the user on a third interface. For instance, the third input could be an operation that triggers the retrieval of a template from the third interface.

[0170] For example, serverless systems can display information through a UI display device. Figure 21 The interface shown (i.e., the third interface) allows users to view and select the pre-configured type and / or template corresponding to the workload to be deployed from a drop-down menu. The pre-configured type indicates the application type corresponding to the workload, and the options for the pre-configured type may include, but are not limited to, one or more of compute-intensive resources, databases, machine learning (ML) services, etc.; the options for the template may include, but are not limited to, optimized web server, first template, second template, and third template, etc.

[0171] according to Figure 21 The interface shown allows the UI display device to send a template request to the serverless system after the user selects a template (e.g., a second template) to request parameters (e.g., a second parameter) and cost values ​​(e.g., a second cost value) corresponding to the second template. Similarly, according to... Figure 21 As shown in the interface, users can also choose to optimize the web server, the first template, or the third template, or other templates. For example, after the user selects the first template, the UI display device can send a template request to the serverless system to request the first parameter and the first cost value corresponding to the first template. Similarly, after the user selects the third template, the UI display device can send a template request to the serverless system to request the third parameter and the third cost value corresponding to the third template.

[0172] It should be noted that, according to Figure 21 The interface shown allows users to define attributes (such as the speed at which workloads are scaled and / or scheduled, the type of compute resources used for the workload, or the workload's tolerance for interruptions) by clicking [the relevant button]. Figure 21 The virtual "Next" button shown leads to the next level of the interface to define properties. When the user clicks... Figure 21After clicking the virtual "Next" button, the resource configuration interface can be displayed by the UI display device. Furthermore, after the user completes the configuration or adjustment of one or more attributes on the resource configuration interface, the UI display device can send a first parameter corresponding to one or more attributes to the serverless system via a first request.

[0173] For example, the UI display device can send a template request to the serverless system based on a fifth input from the user on the second interface. For instance, the fifth input could be an operation that triggers the retrieval of a template from the second interface.

[0174] For example, serverless systems can display information through a UI display device. Figure 22 The interface shown (i.e., the second interface) allows users to view and select the templates they wish to use via the virtual "Add More" button. Template options may include, but are not limited to, optimized web server, first template, second template, third template, etc.

[0175] according to Figure 22 The interface shown allows the UI display device to send a template request to the serverless system after the user clicks the virtual button "Add More" and selects a template (e.g., the second template), requesting parameters (e.g., as the second parameter) and cost values ​​(e.g., as the second cost value) corresponding to the second template. Similarly, according to... Figure 22 As shown in the interface, after clicking the virtual button "Add More," users can select to optimize the web server, the first template, or the third template, among other templates. For example, after the user selects the first template, the UI display device can send a template request to the serverless system to request the first parameter and the first cost value corresponding to the first template. Similarly, after the user selects the third template, the UI display device can send a template request to the serverless system to request the third parameter and the third cost value corresponding to the third template.

[0176] In some embodiments, the serverless system may send a resource configuration template to the UI display device in response to a template request. Upon receiving the resource configuration template, the UI display device may display the corresponding parameters and cost values ​​on an interface (e.g., a fifth interface), including displaying the corresponding parameters and cost values ​​in the form of attribute parameters and / or views, thereby enhancing the visualization of parameters and costs and improving the user experience.

[0177] For example, if the user selects Figure 21 The second template on the interface shown allows the UI display device to display content based on the resource configuration template from the serverless system. Figure 23The interface shown includes a second parameter and a second cost value corresponding to the second template.

[0178] For example, if the user clicks Figure 22 After selecting "Add More" on the interface shown, choosing the second template allows the UI display device to display content based on the resource configuration template from the serverless system. Figure 24 The interface shown includes a second parameter and a second cost value corresponding to the second template. For example... Figure 24 As shown, the UI display device can also continue to retain the received historical cost values ​​(e.g., the first cost value and the third cost value) for users to compare, so that users can choose the better workload deployment plan.

[0179] It should be noted that, Figure 25 This is just one example of multiple cost values ​​displayed on the UI display device. In some embodiments, such as Figure 25 As shown, after receiving the resource configuration template from the serverless system, the UI display device can display the second parameter and the second cost value corresponding to the resource configuration template on the interface, while simultaneously canceling the display of the earliest received cost value (i.e., the first cost value) and its corresponding parameter (i.e., the first parameter). This embodiment of the invention does not specifically limit the number of cost values ​​that the UI display device can display or the specific mechanism by which the UI display device displays multiple cost values; these can be determined based on the specific device capabilities, usage scenarios, and application settings.

[0180] In some embodiments, the serverless system can also output third-party cost suggestion information so that the UI display device can show this information to the user. For example, Figure 25 This shows an example of a third-party cost suggestion information displayed to a user by a UI display device. Figure 25 The third cost recommendation information shown includes suggestions for adopting a workload deployment scheme corresponding to the first cost value (e.g., $12 per replica per hour).

[0181] In some embodiments, the third cost recommendation information can also be used to indicate the impact of the difference between the second parameter and at least one of the first or third parameters on the cost value of the resources required to deploy the workload. For example... Figure 25 As shown, adjusting the guaranteed copy rate to 30% reduces the unit time cost by $4, allowing users to visually see the impact of different attribute sets on costs.

[0182] In some embodiments, the third cost recommendation information may also include recommendations for adjusting attributes, such as a recommendation to adjust an attribute to a certain parameter.

[0183] In S2002, after receiving the fourth feedback on the second cost value, the serverless system configures the resources for deploying workloads according to the second parameter.

[0184] For example, the fourth feedback on the second cost value can determine that the resources used to deploy the workload are configured according to the second parameter corresponding to the second cost value.

[0185] In some embodiments, for example, the fourth feedback may be a configuration request that requests the serverless system to configure resources for deploying workloads according to the second parameter. The configuration request may include the second parameter and / or an identifier for a resource configuration template. Using the identifier, the serverless system can quickly identify a template containing information about the resources on which the user wants to deploy workloads.

[0186] For example, if the user is satisfied with the second cost value, that is, if the user can afford the cost corresponding to the second cost value and decides to adopt the workload deployment plan, the user can trigger the UI display device to send a fourth feedback on the second cost value to the serverless system.

[0187] For example, when the UI display device receives a user click Figure 23 , Figure 24 or Figure 25 After clicking the "Confirm Deployment" virtual button on the interface, the UI display device can send a fourth feedback on the second cost value to the configuration manager in the serverless system. Upon receiving this fourth feedback on the second cost value, the configuration manager in the serverless system can instruct the scheduler to configure the resources used to deploy the workload according to the second parameter.

[0188] Of course, in some embodiments of the present invention, if the user is not satisfied with the second cost value, that is, if the user believes that the cost corresponding to the second cost value is too high and decides to abandon the workload deployment plan, the user can also adjust the attributes again to trigger the UI display device to send a fifth feedback on the second cost value to the serverless system. Then, the serverless system can estimate the cost value again based on the adjusted attribute parameters and output the cost value, which will not be elaborated here.

[0189] Alternatively, if the user, by comparing the second cost value with at least one of the first or third cost values, ultimately decides to adopt the workload deployment scheme corresponding to the first cost value, the user can adjust the attribute parameters back to the attribute parameter set corresponding to the previous first cost value. Alternatively, the user can directly adjust the attribute parameters back to the attribute parameter set corresponding to the previous first cost value by clicking the shortcut button provided on the resource configuration interface. This embodiment of the invention does not limit the specific implementation methods and processes.

[0190] Alternatively, if a user, by comparing the second cost value with at least one of the first or third cost values, ultimately decides to adopt the workload deployment scheme corresponding to the third cost value, the user can adjust the attribute parameters back to the attribute parameter set corresponding to the previous third cost value. Alternatively, the user can directly adjust the attribute parameters back to the attribute parameter set corresponding to the previous third cost value by clicking the shortcut button provided on the resource configuration interface. This embodiment of the invention does not limit the specific implementation methods and processes.

[0191] In some embodiments, Figure 20 The method shown can be implemented independently. This allows users to directly select a template from the existing options and, based on the parameters and cost values ​​in the template sent by the serverless system, decide whether to adopt the workload deployment scheme corresponding to that template. This method greatly simplifies user operations, improves user experience, and reduces the computational process of the serverless system.

[0192] In some embodiments, Figure 20 The steps in the method shown can be executed Figure 5 , Figure 9 or Figure 17 The steps in the method shown are executed afterward. That is, users can select a template from the existing templates, compare the cost of that template with the cost of other workload deployment options, and thus determine the workload deployment option to use. This method helps users make cost trade-offs and workload deployment decisions, thereby improving the user experience.

[0193] For example, Figure 26 This is a flowchart of another resource configuration method provided in an embodiment of the present invention.

[0194] like Figure 26 As shown, before executing S501 to S504, the resource configuration method provided in this embodiment of the invention may further include step S2601.

[0195] In S2601, the serverless system responds to a template request by outputting a resource configuration template.

[0196] The resource configuration template includes a second parameter and a second cost value. The first parameter is a parameter obtained by adjusting the second parameter. The second parameter includes a set of one or more of the following second attribute parameters: the speed at which the workload is scaled and / or scheduled, the type of compute resources used for the workload, or the workload's tolerance to interruptions.

[0197] In some embodiments, the serverless system can send a resource configuration template to the UI display device in response to a template request. After receiving the resource configuration template, the UI display device can display the corresponding parameters and cost values ​​on the interface.

[0198] For example, if the user selects Figure 21 The second template on the interface shown allows the UI display device to display content based on the resource configuration template from the serverless system. Figure 23 The interface shown includes a second parameter and a second cost value corresponding to the second template.

[0199] For a detailed introduction to S2601, please refer to the introduction of S2001 above; it will not be repeated here.

[0200] In S501, the serverless system receives the first request.

[0201] The first request includes a first set of parameters, which includes one or more of the following first attribute parameters: the speed at which the workload to be deployed is scaled and / or scheduled, the type of compute resources used by the workload, and the workload’s tolerance to interruptions.

[0202] In some embodiments, the first request is sent by the UI display device to the serverless system. For example, the UI display device may send a set of parameters (e.g., a first parameter) corresponding to one or more attributes (e.g., the speed at which the workload is scaled and / or scheduled, the type of computing resources used for the workload, or the workload's tolerance to interruptions) to the serverless system via the first request, based on adjustments made by the user on one or more attributes (e.g., the speed at which the workload is scaled and / or scheduled, the type of computing resources used for the workload, or the workload's tolerance to interruptions) in a resource configuration interface.

[0203] For example, the interface displayed by the UI display device after the serverless system outputs the resource configuration template is as follows: Figure 23 As shown. Responding to the user's... Figure 23 The interface shown will ensure that the replica count is adjusted to 50%, and the UI display device can send a first request to the serverless system. The first request includes a set of one or more adjusted attribute parameters, i.e., the first parameter.

[0204] For example, the interface displayed by the UI display device after the serverless system outputs the resource configuration template is as follows: Figure 23 As shown. Responding to the user's... Figure 23 On the interface shown, by dragging the point of the guaranteed replica item in the first view corresponding to the first parameter, the guaranteed replica item is adjusted to 50%. The UI display device can then send a first request to the serverless system. The first request includes a set of one or more adjusted attribute parameters, i.e., the first parameter.

[0205] In S502, the serverless system estimates a first cost value based on a first parameter.

[0206] For example, in response to the user Figure 23 The interface shown will ensure that the copy item is adjusted to 50%, or respond to the user's actions. Figure 23 On the interface shown, by dragging the point of the guaranteed replica item in the first view corresponding to the first parameter, the guaranteed replica item is adjusted to 50%. The UI display device can then send a first request to the serverless system. Furthermore, the serverless system can estimate a first cost value based on the adjusted guaranteed replica item and other parameters.

[0207] In S503, the serverless system outputs the first cost value.

[0208] For example, in response to the user Figure 23 The interface shown will ensure that the copy item is adjusted to 50%, or respond to the user's actions. Figure 23 On the interface shown, by dragging the point of the guaranteed replica item in the first view corresponding to the first parameter, the guaranteed replica item can be adjusted to 50%. This allows the serverless system to... Figure 27 The interface shown (i.e., the second interface) displays a first cost value to the user, for example, $12 per copy per hour.

[0209] It should be noted that, in the embodiments of the present invention, only examples are given to illustrate the use of... Figure 23 Dragging a corresponding point on a line in the middle view toward the geometric center point away from the view will cause the value of the corresponding configuration item to increase. However, the embodiments of the present invention are not limited to this. In some other embodiments, dragging the point toward the geometric center of the view will cause the value of the corresponding configuration item to decrease.

[0210] In some embodiments, such as Figure 27 As shown, the second interface may also include a first view and a second view. The first view can display the first parameter in geometric shape according to the user's configuration or adjustment operations on one or more attributes, and the second view can display the second parameter in geometric shape according to the user's configuration or adjustment operations on one or more attributes, thereby enhancing the visualization of the parameters and improving the user experience.

[0211] In some embodiments, the serverless system may also display first cost recommendation information to the user via a UI display device. For example, the first cost recommendation information may include a suggestion to adopt a workload deployment scheme corresponding to a first cost value.

[0212] In some embodiments, the first cost recommendation information can also be used to indicate the impact of the difference between the first and second parameters on the cost of the resources required to deploy the workload. For example, by adjusting the guaranteed replication item to 50%, the cost per unit time increases by $2, allowing users to visually see the impact of different attribute combinations on costs.

[0213] In some embodiments, the first cost recommendation information may also include recommendations for adjusting attributes, such as adjusting an attribute to a certain parameter.

[0214] In S504, after receiving the first feedback on the first cost value, the serverless system configures resources for deploying workloads according to the first parameter.

[0215] For example, if a user is satisfied with the first cost value—that is, if the user can afford the cost corresponding to the first cost value and decides to adopt the workload deployment plan—the user can trigger the UI display device to send the first feedback on the first cost value to the serverless system.

[0216] In some embodiments of the present invention, if a user is dissatisfied with the first cost value—that is, if the user believes the cost corresponding to the first cost value is too high and decides to abandon the workload deployment plan—the user can adjust the attribute parameters on the resource configuration interface displayed on the UI display device to obtain an ideal workload deployment plan that meets the user's cost expectations. In this way, the UI display device can send a second feedback on the first cost value to the serverless system based on the user's adjustment of the attribute parameters. The second feedback includes a third parameter, which the serverless system can use to estimate a third cost value and display it to the user.

[0217] Alternatively, if a user compares the second cost value with the first cost value and ultimately decides to adopt the workload deployment scheme corresponding to the second cost value, the user can adjust the attribute parameters back to the attribute parameter set corresponding to the previous second cost value. Alternatively, the user can directly adjust the attribute parameters back to the attribute parameter set corresponding to the previous second cost value by clicking the shortcut button provided in the resource configuration interface. This embodiment of the invention does not limit the specific implementation methods and processes.

[0218] It should be noted that the following description only uses the above-described embodiments of the present invention as examples, in which the user configures workload attributes or resource scaling attributes through a UI interface (e.g., the first interface, the second interface, the third interface, the fourth interface, the fifth interface, and other resource configuration interfaces), while cost values, cost suggestion information, and other information are displayed to the user on the UI interface; however, the embodiments of the present invention are not limited to the methods or approaches for configuring or adjusting scaling / scheduling preferences, computing resource types, or guaranteed replication attributes, nor are they limited to the methods or approaches for displaying cost values, cost suggestion information, and other information to the user.

[0219] For example, in some embodiments, the serverless system may also provide a programmatic interface through which users can configure or adjust scaling / scheduling preferences, compute resource types, or attribute parameters of guaranteed replication. Additionally, the serverless system may also display cost values, cost recommendations, and other information on a corresponding interface via the programmatic interface.

[0220] For example, referring to Table 1 below, Table 1 shows some examples of programming instructions provided in embodiments of the present invention.

[0221] Table 1

[0222] It should be understood that the various solutions of the embodiments of the present invention can be reasonably combined and used, and the explanations or descriptions of various terms appearing in the embodiments can be mutually referenced or explained in the various embodiments. The embodiments of the present invention are not limited thereto.

[0223] It should also be understood that, in the various embodiments of the present invention, the numerical values ​​of the sequence numbers of the above processes do not indicate the execution order. The execution order of each process is determined according to its function and inherent logic, and should not constitute any limitation on the various processes of the embodiments of the present invention.

[0224] It is understood that, in order to achieve the functionality of any of the above embodiments, the serverless system or its various modules include hardware structures and / or software modules for performing the corresponding functions. It will be apparent to those skilled in the art that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a particular function is executed in hardware or by computer software driving hardware execution depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but these implementations should not be considered beyond the scope of the present invention.

[0225] In this embodiment of the invention, the serverless system or its various modules can be divided into functional modules. For example, they can be divided into functional modules corresponding to various functions, or two or more functions can be integrated into a single processing module. These integrated modules can be implemented in hardware or as software functional modules. It should be noted that the module division in this embodiment is illustrative and only represents a logical functional division; other division methods may exist in actual implementation.

[0226] For example, refer to Figure 28 This invention provides a serverless system or a resource configuration device, comprising a transceiver, a memory, and a processor. Here, the transceiver supports the system or device in sending and receiving communication signals; the memory supports the system or device in storing computer program instructions; and the processor supports the system or device in executing the computer program instructions, causing the device to implement the resource configuration method provided in any of the above embodiments.

[0227] For example, refer to Figure 29 This invention provides a terminal device including a transceiver, a memory, and a processor. Here, the transceiver supports the terminal device in sending and receiving communication signals; the memory supports the terminal device in storing computer program instructions; and the processor supports the terminal device in executing the computer program instructions to enable the terminal device to implement the resource configuration method provided in any of the above embodiments.

[0228] It should also be understood that serverless systems or their modules can be implemented in software and / or hardware, without limitation. In other words, devices (e.g., terminal devices or proxy servers) are presented as functional modules. The term "module" as used herein can refer to an application-specific integrated circuit (ASIC), a processor and memory for executing one or more software or firmware programs, integrated logic circuits, and / or other devices capable of providing the aforementioned functions.

[0229] In an alternative approach, when using software to implement data transmission, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program is loaded and executed by a computer, it can implement all or part of the processes or functions described in the embodiments of the present invention. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or any other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic cable, or digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, or microwave) means. The computer-readable storage medium can be any available medium accessible to a computer, or it can be a data storage device, such as a server, data center, etc., integrating one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, magnetic disk, or magnetic tape), an optical medium (e.g., digital versatile disk (DVD)), a semiconductor medium (e.g., solid-state drive (SSD)), etc.

[0230] The steps of the methods or algorithms described in the embodiments of this invention can be implemented in hardware or by a processor executing software instructions. The software instructions may include corresponding software modules, which may be stored in random access memory (RAM), flash memory, read-only memory (ROM), erasable programmable read-only memory (EPROM), electrically erasable read-only memory (EEPROM), registers, hard disks, portable hard disks, read-only optical disk drives (CD-ROMs), or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be part of the processor. The processor and storage medium may be within an application-specific integrated circuit (ASIC). Additionally, the ASIC may be located in a serverless system or in various modules therein. Of course, the processor and storage medium may also exist as discrete components.

[0231] Through the description of the above embodiments, those skilled in the art will clearly recognize that, for the sake of convenience and brevity, the illustrative examples are based solely on the functional module division described above. In practical applications, the above functions can be assigned to different functional devices as needed. That is, the internal structure of the device is divided into different functional modules to complete all or part of the above functions.

[0232] The above description is merely a specific implementation of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions within the scope of the present invention can be readily conceived by those skilled in the art, and such variations or substitutions should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A resource allocation method, characterized in that, The method is executed in a serverless system and includes: Receive a first request including a first parameter, wherein the first parameter includes a set of one or more of the following first attribute parameters: the speed at which the workload to be deployed is scaled and / or scheduled, the type of computing resources of the workload, or the workload's tolerance to interruption. Estimate the first cost value based on the first parameter; Output the first cost value; After receiving the first feedback on the first cost value, resources for deploying the workload are configured according to the first parameter.

2. The method according to claim 1, characterized in that, The method further includes: before receiving the first request, In response to the template request, a resource configuration template including a second parameter and a second cost value is output, wherein, The first parameter is obtained by adjusting the second parameter, which includes a set of second attribute parameters including one or more of the following: the speed at which the workload is scaled and / or scheduled, the type of computing resources of the workload, and the workload's tolerance to interruptions.

3. The method according to claim 2, characterized in that, The method further includes: Output first cost recommendation information, wherein the first cost recommendation information is used to indicate the impact of the difference between the first parameter and the second parameter on the cost value of the resources required to deploy the workload.

4. The method according to any one of claims 1 to 3, characterized in that, The attribute parameters for scaling and / or scheduling the workload include the proportion of the workload to be deployed on at least one type of node among hot pool nodes, cold pool nodes, or instant pool nodes, wherein... The scaling and / or scheduling of the hot pool nodes is faster than the scaling and / or scheduling of the cold pool nodes, and the scaling and / or scheduling of the cold pool nodes is faster than the scaling and / or scheduling of the instantaneous pool nodes.

5. The method according to any one of claims 1 to 4, characterized in that, The attribute parameters of the computing resource type of the workload include at least one of input / output (I / O) intensive resources, central processing unit (CPU) intensive resources, or memory intensive resources.

6. The method according to any one of claims 1 to 5, characterized in that, The attribute parameters of the workload's tolerance to interruptions include: The proportion of the workload to be deployed on on-demand instances, and / or The proportion of the workload to be deployed on the spot instance.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: In response to a second feedback on the first cost value, a third cost value is estimated, wherein the second feedback includes a third parameter, the third parameter being a set of one or more of the following third attribute parameters: the speed at which the workload is scaled and / or scheduled, the type of computing resources of the workload, or the workload’s tolerance to interruptions. Output the third cost value; After receiving third feedback on the third cost value, resources for deploying the workload are configured according to the third parameter.

8. The method according to claim 7, characterized in that, The method further includes: Output second cost recommendation information, wherein the second cost recommendation information is used to indicate the impact of the difference between the first parameter and the third parameter on the cost value of the resources required to deploy the workload.

9. A resource allocation method, characterized in that, The method is executed in a terminal and includes: In response to a user’s first input on a first interface of the terminal, a first request including a first parameter is sent to the serverless system, wherein the first parameter includes a set of one or more first attribute parameters, namely, the speed at which the workload to be deployed is scaled and / or scheduled, the type of computing resources of the workload, or the workload’s tolerance to interruption. Receive from the serverless system a first cost value for the resources used to deploy the workload, estimated based on the first parameter; Display a second interface including the first cost value; In response to the user's second input on the second interface, a first feedback is sent to the serverless system, wherein the first feedback is used to instruct the resources for deploying the workload to be configured according to the first parameter.

10. The method according to claim 9, characterized in that, The method further includes: before sending the first request including the first parameter to the serverless system in response to the first input by the user on the first interface of the terminal, In response to the user's third input on the third interface, a template request is sent to the serverless system; The serverless system receives a first resource configuration template including a second parameter and a second cost value, wherein the second parameter includes a set of second attribute parameters of one or more of the speed at which the workload is scaled and / or scheduled, the type of computing resources of the workload, or the tolerance of the workload to interruption, and the second cost value is a cost value of resources for deploying the workload estimated based on the second parameter. The first interface displays the second parameter and the second cost value, wherein the first parameter is a parameter obtained by adjusting the second parameter.

11. The method according to claim 10, characterized in that, The second interface further includes a first view and a second view, wherein the first view is used to display the first parameter in geometric form, and the second view is used to display the second parameter in geometric form.

12. The method according to claim 10 or 11, characterized in that, The method further includes: The second interface receives first cost recommendation information from the server, wherein the first cost recommendation information is used to indicate the impact of the difference between the first parameter and the second parameter on the cost value of the resources required to deploy the workload, and the second interface also includes the first cost recommendation information.

13. The method according to any one of claims 9 to 12, characterized in that, The method further includes: In response to the user's fourth input on the second interface, a second feedback including a third parameter is sent to the serverless system, wherein the third parameter is a parameter obtained by adjusting the first parameter, and the third parameter includes a set of third attribute parameters of one or more of the speed of scaling and / or scheduling the workload, the computing resource type of the workload, or the tolerance of the workload to interruption, and the second feedback is used to indicate a third cost value of the resources required to deploy the workload based on the third parameter; Receive the third cost value from the serverless system; A fourth interface is displayed, including the third cost value.

14. The method according to claim 13, characterized in that, The fourth interface further includes a first view and a third view, wherein the first view is used to display the first parameter in geometric form, and the third view is used to display the third parameter in geometric form.

15. The method according to claim 13 or 14, characterized in that, The method further includes: The fourth interface receives second cost recommendation information from the serverless interface, wherein the second cost recommendation information is used to indicate the impact of the difference between the first parameter and the third parameter on the cost value of the resources required to deploy the workload, and the fourth interface also includes the second cost recommendation information.

16. The method according to any one of claims 13 to 15, characterized in that, The method further includes: before sending the second feedback to the serverless system. Receive the fourth input, wherein receiving the fourth input includes: Receive adjustments to the first view, or Receive one or more of the following attribute parameters input on the second interface: the speed of scaling and / or scheduling the workload, the computing resource type of the workload, or the tolerance of the workload to interruption.

17. The method according to any one of claims 9 to 16, characterized in that, The method further includes: In response to the user's fifth input on the second interface, a second resource configuration template including a fourth parameter and a fourth cost value is received from the serverless system, wherein the fourth parameter includes a set of fourth attribute parameters of one or more of the speed at which the workload is scaled and / or scheduled, the type of computing resources of the workload, or the tolerance of the workload to interruption, and the fourth cost value is a cost value of resources for deploying the workload estimated based on the fourth parameter; The display includes a fifth interface comprising a first view and a fourth view, wherein the first view is used to display the first parameter in geometric form, and the fourth view is used to display the fourth parameter in geometric form.

18. The method according to any one of claims 9 to 17, characterized in that, The method further includes: In response to the user's sixth input on the second interface, a template save request is sent to the serverless system to save the first parameter and the first cost value as a third resource configuration template.

19. A resource allocation method, characterized in that, The method is executed in a serverless system and includes: In response to a template request from a terminal, a first resource configuration template is sent to the terminal. The first resource configuration template includes a second parameter and a second cost value. The second parameter includes a set of second attribute parameters, which are one or more of the following: the speed at which the workload to be deployed is scaled and / or scheduled, the type of computing resources of the workload, or the workload's tolerance to interruption. The second cost value is the cost of resources used to deploy the workload, estimated based on the second parameter. In response to the configuration request from the terminal, resources for deploying the workload are configured according to the second parameter.

20. The method according to claim 19, characterized in that, The configuration request includes: The second parameter or The identifier of the first resource configuration template.

21. A resource allocation method, characterized in that, The method is executed in a terminal and includes: Send a template request to the serverless system; The serverless system receives a first resource configuration template including a second parameter and a second cost value, wherein the second parameter includes a set of one or more second attribute parameters, such as the speed at which the workload to be deployed is scaled and / or scheduled, the type of computing resources of the workload, or the workload's tolerance to interruption, and the second cost value is the cost of resources used to deploy the workload estimated based on the second parameter. In response to user input, a configuration request is sent to the serverless system, wherein the configuration request is used to request the deployment of resources for the workload according to the second parameter.

22. A resource allocation device, characterized in that, The resource allocation device includes: A transceiver is used to send and receive communication signals; Memory is used to store computer program instructions; A processor for executing the computer program instructions to cause the apparatus to implement the method according to any one of claims 1 to 8 or claims 19 or 20.

23. A serverless system, characterized in that, The serverless system includes: A transceiver is used to send and receive communication signals; Memory is used to store computer program instructions; A processor for executing the computer program instructions to cause the terminal device to implement the method according to any one of claims 1 to 8 or claims 19 or 20.

24. A terminal device, characterized in that, The terminal device includes: A transceiver is used to send and receive communication signals; Memory is used to store computer program instructions; A processor for executing the computer program instructions to cause the apparatus to perform the method according to any one of claims 9 to 18 or claim 21.

25. A resource allocation system, characterized in that, The resource allocation system include: The resource allocation device according to claim 22 or the serverless system according to claim 23; The terminal device according to claim 24.

26. A computer-readable storage medium storing computer program instructions, characterized in that, When executed by processing circuitry in a computer, the computer program instructions cause the computer to implement the method according to any one of claims 1 to 21.

27. A computer program product comprising a command, characterized in that, When executed by a computer, the instructions cause the computer to implement the method according to any one of claims 1 to 21.

28. A chip system, characterized in that, The chip system includes a processing circuit and a storage medium, wherein the storage medium stores computer program instructions, which, when executed by the processing circuit, cause the chip system to implement the method according to any one of claims 1 to 21.

Images