Business Service Colocation via Hierarchical Resource Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing business service colocation technologies face challenges in efficiently co-locating multiple types of services due to resource contention and interference, leading to low resource utilization and stability issues in data centers.
Innovation Solution
A method and system for colocation of multiple business services that determine a target resource level and schedulable resources for each service, dynamically adjusting resource allocation based on deployment conditions to optimize resource utilization and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple types of business services are co-located on a device, then device resource utilization is improved, but resource contention and interference occur between services
Solution Approach 1:
The patent segments resources into different levels (first resource level and second resource level) with different guaranteed resource quantities. High-priority services are allocated resources at the first resource level with higher guaranteed quantities, while low-priority services are allocated at the second resource level with lower guaranteed quantities. This segmentation allows multiple services to co-locate on the same device while preventing resource contention through hierarchical resource isolation.
Solution Approach 2:
The patent applies local quality by providing different guaranteed resource quantities to different services based on their priority levels. Instead of uniform resource allocation, each service receives tailored resource guarantees appropriate to its priority, enabling high-priority services to maintain stability while low-priority services utilize remaining resources, thus resolving the contradiction between resource utilization and service stability.
2Productivity
If existing business service colocation technology is used, then resource utilization improves, but it is not easy to co-locate a plurality of types of business services
Solution Approach 1:
The patent establishes a universal resource scheduling system that can handle multiple types of business services through a standardized multi-level resource allocation mechanism. The system defines resource levels and guaranteed resource quantities that can accommodate various service types (e.g., offline businesses, online businesses, real-time services, non-real-time services) within a unified framework, enabling versatile co-location while maintaining resource utilization efficiency.
3Productivity
If resources are allocated to multiple business services, then resource utilization improves, but resource interference occurs between services
Solution Approach 1:
The patent segments the resource pool into multiple hierarchical levels with different guaranteed resource quantities. By dividing resources into first resource level (for high-priority services) and second resource level (for low-priority services), the system prevents resource interference through hierarchical isolation. Each service operates within its designated resource level, eliminating contention and interference while maintaining high overall resource utilization.
Solution Approach 2:
The patent introduces resource levels and guaranteed resource quantity allocations as intermediary mechanisms between the physical resources and multiple business services. These intermediaries mediate resource access by enforcing priority-based allocation rules, preventing direct resource interference between services while enabling efficient resource utilization through structured resource management.
Data Source
AI summary
Embodiments of this specification disclose methods and systems for colocation of a plurality of business services. The method includes: for one of the business services, determining a target resource level corresponding to the business service; obtaining a target schedulable resource of each of one or more processing devices, where the target schedulable resource is a schedulable resource corresponding to the target resource level; and determining, based on whether the target schedulable resource satisfies a deployment condition corresponding to the business service, a method for deploying the business service on the one or more processing devices.


