Resource Allocation in Distributed Subserver Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing resource allocation mechanisms in computer systems face inefficiencies, particularly in distributed environments, where quota management often leads to excessive resource consumption and slow processing due to reliance on global quotas and inadequate local quota utilization.
Innovation Solution
A method and apparatus that determine resource quotas based on a target subserver's local availability and globally available quotas, allocating resources efficiently by comparing required quotas with local limits and updating quotas dynamically to optimize resource utilization and reduce global quota requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If global quota management is used, then resource consumption can be controlled at a system level, but processing speed decreases and resource utilization becomes inefficient
Solution Approach 1:
The patent segments the global quota into local quotas for each subserver. Each subserver maintains its own quota independently, allowing local resource allocation decisions without waiting for global quota checks. This segmentation enables parallel processing across multiple subservers, significantly improving processing speed while maintaining overall resource consumption control through the global quota framework.
Solution Approach 2:
The patent implements local quota management where each subserver has its own quota parameters and allocation policies. This allows each subserver to make independent resource allocation decisions based on local needs, eliminating the bottleneck of centralized global quota management and improving processing speed while maintaining system-wide resource control.
2Productivity
If local quota management is used, then processing speed increases, but resource wastage occurs due to inadequate local quota utilization
Solution Approach 1:
The patent implements a feedback mechanism where subservers report their quota usage status to the central controller. The central controller uses this feedback information to dynamically adjust global quota allocations and optimize local quota distributions. This feedback loop prevents resource wastage by ensuring that quotas are allocated efficiently based on actual consumption patterns while maintaining fast local processing decisions.
Solution Approach 2:
The patent makes quota management dynamic by allowing real-time adjustment of both global and local quotas based on actual usage patterns. The system can dynamically allocate more quotas to subservers with high demand and reduce quotas from subservers with low usage, preventing resource wastage while maintaining high processing speed through adaptive quota management.
3Reliability
If continuous global quota deduction is performed, then resource allocation is controlled centrally, but system complexity increases and processing becomes slow
Solution Approach 1:
The patent segments the centralized quota management into distributed local quota management. Each subserver independently manages its own quota, eliminating the need for continuous centralized deduction operations. This segmentation reduces system complexity by distributing management responsibilities while maintaining centralized control through periodic synchronization and feedback mechanisms.
Solution Approach 2:
The patent implements preliminary action by pre-allocating quotas to subservers before actual resource requests occur. The central controller can pre-distribute quotas based on predicted needs or historical data, allowing subservers to operate independently without continuous centralized intervention. This reduces system complexity by enabling autonomous local operations while maintaining overall control through periodic adjustments.
4Measurement precision
If frequent global quota requests are made, then resource allocation accuracy is maintained, but processing time increases
Solution Approach 1:
The patent segments frequent global quota requests into local quota operations. Each subserver performs its own quota management locally, eliminating the need for frequent centralized requests. This segmentation maintains resource allocation accuracy through local quota tracking while significantly reducing processing time by enabling autonomous local decisions without waiting for centralized response.
Solution Approach 2:
The patent implements preliminary quota allocation where the central controller pre-distributes quotas to subservers based on predicted needs or historical patterns. This preliminary action reduces the frequency of subsequent quota requests, maintaining allocation accuracy through periodic adjustments while minimizing processing time by reducing the number of communication cycles between central controller and subservers.
Data Source
AI summary
A method and an apparatus for allocating a resource are provided according to the embodiments. The method may include: extracting a resource category matching a target subserver and a required quota corresponding to the resource category from a request sent by a client side; determining a number of subservers is a subserver group receiving the request; determining a globally available quota and a total limit matching the resource category; determining a limit matching the resource category in the target subserver based on the total limit and the number of subservers; and performing resource allocation for a resource quantity corresponding to the required quota of the request based on a comparison between the required quota and the limit of the target subserver. The embodiment achieves reducing the number of applications for the globally available quota, thus accelerating processing the request by the server.