A heterogeneous cloud platform computing power calling budget determination and interception method and system

Abstract

The present application belongs to the technical field of cloud computing and AI computing power management, in particular to a heterogeneous cloud platform computing power calling budget judgment and interception method and system. In view of the existing lack of cross-platform unified budget real-time judgment, heterogeneous billing standardization and consumption authenticity verification mechanism, which leads to uncontrollable cost control and untrustworthy audit data, the following scheme is proposed, which includes the following steps: S1: Establish a unified computing power budget account independent of each cloud platform, and bind the enterprise's AI service account on each cloud platform to the account; S2: Set the total unified budget and the budget allocation quota of each platform. The present application realizes real-time budget control and automatic quota allocation of cross-platform computing power calling by constructing an independent budget account and a unified interception point; with the help of data standardization and interception log comparison, the consumption authenticity is verified, the trusted audit data is provided, and the efficiency, accuracy and reliability of multi-cloud cost management are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of cloud computing and AI computing power management technology, and in particular to a method and system for determining and intercepting computing power call budgets on heterogeneous cloud platforms. Background Technology

[0002] Enterprises typically use multiple cloud platforms simultaneously in their AI computing power consumption, but existing technologies face three technical challenges: 1. Heterogeneous platforms cannot unify budgets. Budget management tools on different cloud platforms are independent of each other, making it impossible for enterprises to view and manage cross-platform budgets in real time within a unified account. When a department runs out of budget on platform A but still has a surplus on platform B, automatic allocation is not possible, leading to wasted resources or business interruptions. 2. Cross-platform billing data cannot be standardized. Billing field names, billing granularity, and timestamp formats differ across cloud platforms. For example, Alibaba Cloud uses resource instance IDs and payable amounts, while Volcano Engine uses application IDs and token consumption counts. Independent auditing systems cannot obtain consistent consumption data, resulting in extremely low reliability of audit conclusions. 3. Lack of cross-platform data verification mechanisms. Existing cloud cost management tools can only aggregate and statistically analyze invoices that have already been generated, and cannot verify whether each expense record reported by the cloud platform corresponds to a real, budget-approved call. This makes cross-platform auditing lack a reliable data foundation.

[0003] Existing technologies lack cross-platform unified budget real-time determination, heterogeneous billing standardization, and consumption authenticity verification mechanisms, resulting in uncontrollable cost control and unreliable audit data. To address these issues, we propose a method and system for determining and intercepting computing power call budgets on heterogeneous cloud platforms. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies, such as the lack of cross-platform unified budget real-time determination, heterogeneous billing standardization, and consumption authenticity verification mechanisms, which lead to uncontrollable cost control and unreliable audit data. Therefore, this invention proposes a method and system for determining and intercepting computing power call budgets on heterogeneous cloud platforms.

[0005] This application provides a method and system for determining and intercepting computing power allocation budgets in heterogeneous cloud platforms, which adopts the following technical solution: A method for determining and intercepting computing power allocation budgets on heterogeneous cloud platforms includes the following steps: S1: Establish a unified computing power budget account independent of each cloud platform, and bind the enterprise's AI service accounts on each cloud platform to this account; S2: Set a unified total budget and budget allocation for each platform, and support dynamic adjustments; S3: Set up a unified interception point on the necessary path of all AI call requests sent to various cloud platforms; S4: When an AI call request reaches the unified interception point, query the remaining budget of the target cloud platform in the unified budget account. If there is enough, allow the request and deduct the amount. If there is not enough, perform cross-platform budget allocation or execute interception. S5: Collect heterogeneous raw billing data from various cloud platforms and map it into a unified standard data model through a data standardization intermediate layer; S6: Compare the standardized consumption data with the interception decision log generated by the unified interception point to perform authenticity verification; S7: Generate a unified audit report across platforms based on verified trusted consumption data.

[0006] Furthermore, in S3, the unified interception point is deployed on the enterprise's local server, an independent third-party management platform, and the API gateway layer provided by the cloud platform.

[0007] Furthermore, in S5, the data standardization middleware layer develops an independent adapter for each cloud platform and presets a field mapping table to unify the heterogeneous field names, timestamp formats, and billing units of each platform.

[0008] Furthermore, in S4, cross-platform budget allocation is carried out by allocating quotas from the remaining budgets of other platforms according to a preset priority algorithm. The allocation rules are preset by the administrator, and the allocation action is automatically executed by the system.

[0009] Furthermore, in S2, the budget determination is achieved through local cached budget data and memory calculation, and the consumption data of each cloud platform is asynchronously synchronized in the background.

[0010] Furthermore, it also includes: based on verified and reliable consumption data, before cross-platform transactions are executed, all stakeholders preset and confirm the revenue sharing rules, and the multi-party revenue sharing is automatically completed according to the preset rules during transaction settlement.

[0011] This invention also proposes a heterogeneous cloud platform computing power call budget determination and interception system to implement the above-mentioned heterogeneous cloud platform computing power call budget determination and interception method, including: a unified account module, a budget allocation module, a unified interception module, a consumption data collection and standardization module, an authenticity verification module, and an audit report generation module.

[0012] In summary, this application includes at least one of the following beneficial technical effects: 1. This solution achieves unified, real-time budget control over computing power calls from multiple heterogeneous cloud platforms by constructing a unified budget account and interception point independent of each cloud platform. It supports automatic quota allocation and avoids resource waste and business interruption. 2. This solution solves the problem of heterogeneous billing data structures across different cloud platforms by using a data standardization middleware layer, providing a consistent and reliable data source for subsequent auditing, analysis, and AI model training cost accounting; 3. This solution compares the interception decision log with the standardized bill, thereby verifying the authenticity of each consumption record from a technical perspective. It can effectively identify abnormal billing, significantly improve the reliability of cross-platform cost auditing, and solve the fundamental problem of lacking a reliable data foundation for cross-platform auditing.

[0013] This invention enables real-time budget control and automatic quota allocation for cross-platform computing power calls by constructing independent budget accounts and unified interception points; it verifies the authenticity of consumption and provides reliable audit data by using data standardization and interception log comparison, effectively improving the efficiency, accuracy and reliability of multi-cloud cost management. Attached Figure Description

[0014] Figure 1 This is a diagram illustrating the overall architecture of a heterogeneous cloud platform computing power call budget determination and interception method and system proposed in this invention. Figure 2 This invention presents a flowchart of the budget determination and cross-platform allocation process for a heterogeneous cloud platform computing power allocation budget determination and interception method and system. Figure 3 This invention presents a flowchart of the heterogeneous billing standardization and authenticity verification process for a heterogeneous cloud platform computing power call budget determination and interception method and system. Detailed Implementation

[0015] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0016] Example 1 Reference Figures 1-3 A method for determining and intercepting computing power usage budget on a heterogeneous cloud platform, comprising the following steps: S1: Establish a unified computing power budget account independent of each cloud platform, and bind the enterprise's AI service accounts on each cloud platform to this account; An independent system establishes a unified computing power budget account for enterprise customers. This account is independent of any cloud platform. Enterprise customers allocate the total budget proportionally or in fixed amounts to the various AI cloud platforms they have connected to. For example, if the total monthly budget is 1 million yuan, 300,000 yuan is allocated to Volcano Engine, 500,000 yuan to Alibaba Cloud, and 200,000 yuan to Huawei Cloud. The allocation amount can be dynamically adjusted and takes effect immediately. S2: Set a unified total budget and budget allocation for each platform, and support dynamic adjustments; Set up a unified interception point on the necessary path of all AI computing power call requests (such as API gateway or SDK layer). This interception point is deployed independently of the cloud platform and intercepts all call requests sent to each cloud platform. When an AI request reaches the unified interception point, the following budget determination process is executed: Identify the target cloud platform for the request; Query the real-time remaining budget for the target platform in the unified budget account; If the remaining budget is sufficient, a clearance signal is returned, requesting clearance to the target cloud platform, and the platform's budget is deducted in real time; If the remaining budget is insufficient, determine whether it can be automatically allocated from the remaining quota of other cloud platforms. If it can be allocated, execute the allocation and return a pass; if it cannot be allocated, return a rejection signal, the request is blocked, and a budget shortage notification is returned to the upper layer application. Real-time implementation: Budget account data is updated in real time in the local cache, budget determination is completed through in-memory calculation, and consumption data of each cloud platform is asynchronously synchronized in the background without affecting the real-time determination of the call path.

[0017] S3: Set up a unified interception point on all the necessary paths of AI call requests sent to various cloud platforms. The unified interception point is deployed on the enterprise's local server, an independent third-party management platform, and the API gateway layer provided by the cloud platform. Raw AI-generated consumer billing data is collected from various cloud platforms via API interfaces. After collection, it is processed uniformly through a data standardization intermediate layer, which includes: Adapter module: Develop independent adapters for each cloud platform and be responsible for pulling raw billing data through the official API interfaces of each platform; Field mapping engine: Maps heterogeneous billing fields from various platforms to a unified standard data model. The standard data model includes at least: unified timestamp, unified user identifier, unified model identifier, unified cost unit, and unified project / business tag. S4: When an AI call request reaches the unified interception point, the remaining budget of the target cloud platform in the unified budget account is queried. If it is sufficient, it is allowed and deducted. If it is insufficient, cross-platform budget allocation or interception is performed. Cross-platform budget allocation is based on a preset priority algorithm to allocate quotas from the remaining budgets of other platforms. The allocation rules are preset by the administrator, and the allocation action is automatically executed by the system. Budget determination is achieved through local cached budget data and memory calculation. The consumption data of each cloud platform is asynchronously synchronized through the background. For the standardized consumption data, authenticity verification is performed. The verification method is as follows: the standardized consumption record returned by the cloud platform is compared with the interception decision log of this invention. The interception decision log records each allowed call request. By matching the request log with the consumption bill, it can be verified whether each expense record reported by the cloud platform corresponds to a real call that has been approved by the budget. If an abnormal record that cannot be matched is found, it is marked and prevented from entering the normal settlement process. S5: Collect heterogeneous raw billing data from various cloud platforms and map it into a unified standard data model through a data standardization middleware layer. The data standardization middleware layer develops an independent adapter for each cloud platform and presets a field mapping table to unify the heterogeneous field names, timestamp formats and billing units of each platform. When cross-platform computing power transactions involve the interests of multiple parties, before the transaction is executed, the system requires all stakeholders to pre-confirm the sharing rules and proportions within the system through the sharing confirmation module. After confirmation, the system generates a sharing confirmation voucher. During transaction settlement, the system automatically executes the sharing based on this voucher without manual intervention. If any party fails to complete the confirmation in advance, the transaction cannot be settled normally. S6: Compare the standardized consumption data with the interception decision log generated by the unified interception point to perform authenticity verification; Based on all the above data (standardized bills, interception logs, verification results, and revenue sharing records), a cross-platform "AI Computing Power Cost Special Audit Report" is generated, which includes a cross-platform AI consumption overview, budget utilization rate of each platform, a list of intercepted call records, a list of abnormal consumption records, and details of multi-party revenue sharing and settlement. S7: Based on verified trusted consumption data, generate a unified audit report across platforms. It also includes: based on verified trusted consumption data, before cross-platform transaction execution, all stakeholders preset and confirm the revenue sharing rules, and the multi-party revenue sharing is automatically completed according to the preset rules during transaction settlement.

[0018] This embodiment also proposes a heterogeneous cloud platform computing power call budget determination and interception system to implement the above-mentioned heterogeneous cloud platform computing power call budget determination and interception method, including: a unified account module, a budget allocation module, a unified interception module, a consumption data collection and standardization module, an authenticity verification module, and an audit report generation module.

[0019] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for determining and intercepting computing power allocation budgets in heterogeneous cloud platforms, characterized in that: Includes the following steps: S1: Establish a unified computing power budget account independent of each cloud platform, and bind the enterprise's AI service accounts on each cloud platform to this account; S2: Set a unified total budget and budget allocation for each platform, and support dynamic adjustments; S3: Set up a unified interception point on the necessary path of all AI call requests sent to various cloud platforms; S4: When an AI call request reaches the unified interception point, query the remaining budget of the target cloud platform in the unified budget account. If there is enough, allow the request and deduct the amount. If there is not enough, perform cross-platform budget allocation or execute interception. S5: Collect heterogeneous raw billing data from various cloud platforms and map it into a unified standard data model through a data standardization intermediate layer; S6: Compare the standardized consumption data with the interception decision log generated by the unified interception point to perform authenticity verification; S7: Generate a unified audit report across platforms based on verified trusted consumption data.

2. The method for determining and intercepting computing power allocation budget in a heterogeneous cloud platform according to claim 1, characterized in that: In S3, the unified interception point is deployed on the enterprise's local server, an independent third-party management platform, and the API gateway layer provided by the cloud platform.

3. The method for determining and intercepting computing power allocation budget in a heterogeneous cloud platform according to claim 2, characterized in that: In S5, the data standardization middleware layer develops an independent adapter for each cloud platform and presets a field mapping table to unify the heterogeneous field names, timestamp formats and billing units of each platform.

4. The method for determining and intercepting computing power allocation budget in a heterogeneous cloud platform according to claim 3, characterized in that: In S4, cross-platform budget allocation is carried out by allocating quotas from the remaining budgets of other platforms according to a preset priority algorithm. The allocation rules are preset by the administrator, and the allocation action is automatically executed by the system.

5. The method for determining and intercepting computing power allocation budget in a heterogeneous cloud platform according to claim 4, characterized in that: In S2, the budget determination is achieved through local cached budget data and memory calculation, and the consumption data of each cloud platform is asynchronously synchronized in the background.

6. The method for determining and intercepting computing power allocation budget in a heterogeneous cloud platform according to claim 5, characterized in that: Also includes: Based on verified and reliable consumption data, all stakeholders preset and confirm the revenue sharing rules before cross-platform transactions are executed, and the multi-party revenue sharing is automatically completed according to the preset rules during transaction settlement.

7. A heterogeneous cloud platform computing power call budget determination and interception system, used to implement the heterogeneous cloud platform computing power call budget determination and interception method according to any one of claims 1 to 6, characterized in that: include: The module includes a unified account module, a budget allocation module, a unified interception module, a consumption data collection and standardization module, an authenticity verification module, and an audit report generation module.

Images