Container Migration Selection Using Response-Time Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Container virtualization systems experience performance deterioration due to resource interference among containers, which is not effectively addressed by existing methods, leading to suboptimal scalability and service degradation.
Innovation Solution
A method and program for determining a container to be migrated by identifying containers vulnerable to interference based on resource usage rates and response time increases, using an interference index to select a destination node with minimal interference, thereby reducing performance degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If containers are migrated frequently to balance load, then system scalability is improved, but performance deterioration due to interference increases
Solution Approach 1:
The system performs preliminary assessment of container performance and node interference levels before migration decisions are made. By evaluating response time changes and interference indices in advance, the system可以避免 migrating containers to nodes that would cause or experience performance deterioration, thus maintaining reliability while enabling scalability.
Solution Approach 2:
The system continuously monitors container response times and node resource usage, using this feedback to dynamically adjust migration decisions. The interference index calculation and response time tracking create a closed-loop control system that adapts to changing conditions, allowing the system to scale while maintaining performance through data-driven migration choices.
2Reliability
If containers are migrated to suppress performance deterioration, then container performance is improved, but migration overhead and system complexity increase
Solution Approach 1:
The system enables containers to effectively self-manage their placement by automatically monitoring their own response times and the interference levels of candidate nodes. The container migration decision-making process is autonomous, using pre-calculated interference indices and real-time performance data to determine optimal migration targets without requiring complex external orchestration.
Solution Approach 2:
The system uses quantitative parameters such as response time changes and interference indices to objectively determine migration necessity and targets. By transforming the complex migration decision problem into a parameter-based evaluation system, the approach simplifies migration management while effectively suppressing performance deterioration through data-driven decisions.
3Productivity
If resource usage is increased to improve service capacity, then system productivity is improved, but interference between containers increases causing performance deterioration
Solution Approach 1:
The system dynamically adjusts container placement based on real-time resource usage patterns and interference levels. As resource usage increases to meet service capacity demands, the system continuously recalculates interference indices and response times, dynamically migrating containers to maintain optimal performance. This dynamic adaptation allows the system to scale service capacity while preventing performance deterioration through real-time rebalancing.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for determining a container to be migrated, the method is implemented by a first node executing a plurality of containers, the method including, determining a first container included in the plurality of the containers as the container to be migrated to a second node, wherein a response time of the first container increases in a time zone in which a usage rate of a resource used by the first node increases, and the first container includes a parameter that affects the increase in the response time, where the parameter does not exceed a predetermined value in the time zone.