Dynamic Resource Regulation for FaaS Containers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current FaaS platforms face low resource utilization as they allocate fixed-sized resources to containers without considering actual memory usage, leading to inefficient use of resources.
Innovation Solution
A resource processing method that acquires a target container and its current resource allocation, regulates the resources to match the actual usage of the target function, and runs the function based on these optimized resources, thereby improving resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed-sized resources are allocated to containers, then resource allocation is simple and reliable, but resource utilization is low
Solution Approach 1:
The patent implements dynamic resource regulation by continuously monitoring container memory usage and adjusting allocated resources accordingly. The system transitions from static fixed-size allocation to dynamic flexible allocation, allowing resources to be reallocated based on actual usage patterns, thereby improving resource utilization while maintaining allocation reliability through controlled regulation mechanisms.
Solution Approach 2:
The patent changes the resource allocation parameter from fixed size to variable size based on actual memory usage. By monitoring memory consumption patterns and adjusting the allocated resource parameters dynamically, the system optimizes the balance between allocation reliability and resource utilization efficiency.
2Productivity
If resources are reduced to match actual usage, then resource utilization improves, but function performance may degrade
Solution Approach 1:
The patent implements a feedback mechanism that continuously monitors container performance metrics and adjusts resource allocation based on observed behavior. By feeding back performance data to the resource regulation system, the patent ensures that resources are reduced only when actual usage justifies it, while maintaining performance thresholds through adaptive control.
Solution Approach 2:
The patent performs preliminary resource allocation with some buffer capacity, then gradually regulates resources downward only when usage patterns consistently indicate excess capacity. This preliminary over-allocation ensures performance reliability is maintained before optimization begins, preventing performance degradation during the transition.
Data Source
AI summary
The present disclosure discloses a resource processing method and a storage medium. The method includes: acquiring a target container of a target function, where the target container is configured to run the target function; acquiring a current resource that is already allocated to the target container; regulating the current resource to a target resource of the target container, where the target resource is determined based on the target function; and running, based on the target resource, the target function in the target container. According to the present disclosure, the technical problem of low resource utilization can be solved.