Multi-Site Microservice Placement for Performance-Based Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques struggle to efficiently manage resource allocation and deployment of microservices across multiple site systems, particularly in AI and data analysis, due to the complexity of considering execution environments and resource status differences between sites.
Innovation Solution
An arrangement plan search device that utilizes a storage unit and processor to identify resource amounts required for microservices based on performance targets, determining optimal allocation across multiple site systems using service resource amount performance information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If resource allocation is managed considering execution environment and resource status differences between multiple sites, then arrangement plan accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the resource allocation problem by site system, maintaining separate execution environment information and resource status data for each site. The arrangement plan search device processes each site's microservices independently while considering inter-site dependencies, thereby achieving accurate multi-site resource allocation without overwhelming complexity in a single centralized system.
Solution Approach 2:
The patent introduces an intermediary information structure that bridges the complexity of multi-site differences. This intermediary layer standardizes the representation of execution environments and resource statuses across sites, allowing the system to handle heterogeneous site configurations through a unified processing framework without directly managing each site's unique characteristics.
2Measurement precision
If the number of steps for analyzing data in distributed sites increases, then analysis accuracy is improved, but loss of time increases
Solution Approach 1:
The patent performs preliminary actions by pre-establishing execution environment information and resource status data for multiple site systems before actual data analysis begins. The arrangement plan search device determines optimal microservice placements and resource allocations in advance, so that when data analysis is executed, the system can directly implement pre-computed plans without performing complex calculations during the analysis process itself.
Solution Approach 2:
The patent ensures continuity of useful action by maintaining persistent execution environment information and resource status data structures that are continuously updated as sites change. This allows the system to quickly query current state information during data analysis without re-performing comprehensive site assessments, thereby maintaining analysis accuracy while reducing time consumption through efficient data access.
Data Source
AI summary
In an application platform, an application performance model and a data store performance model in each of a plurality of site systems are stored, and a processor receives target performance information, determines an application allocated resource amount for the plurality of site systems based on the application performance model, determines required performance of a data store for the plurality of site systems, determines a data store allocated resource amount for the plurality of site systems based on the data store performance model, and searches for an arrangement plan of the application and the data store capable of implementing the application allocated resource amount and the data store allocated resource amount in the plurality of site systems.