Container Scaling via Pre-stored Load Configurations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In microservice architecture, increasing the number of containers to reduce response time can take time and may not be sufficient, especially when the load on one container becomes a bottleneck, leading to delayed response times in services like social networking, electronic commerce, and online gaming.
Innovation Solution
A service management program that acquires current input loads and refers to stored container number information to identify the optimal number of containers needed to reduce response times, allowing for immediate scaling by increasing the number of containers corresponding to previously recorded input loads that achieved reduced response times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the number of containers is increased sequentially to reduce response time, then the response time is eventually reduced, but it takes excessive time for the delay to be reduced
Solution Approach 1:
The system performs preliminary actions by storing container number information in advance for various input load levels. When a response time delay occurs, the service management device refers to this pre-stored information to immediately determine the appropriate container number, eliminating the need for sequential trial-and-error scaling and enabling instant response time reduction.
2Loss of time
If the number of containers is increased to reduce response time, then response time is reduced, but resource overallocation occurs
Solution Approach 1:
The system changes parameters by storing and referencing specific container number configurations corresponding to different input load levels. Instead of blindly increasing container numbers, the service management device selects the precise container count from stored configurations that matches the current input load, achieving response time reduction without resource overallocation.
3Productivity
If sequential scaling is used to increase container numbers, then the service can handle increased load, but the process is inefficient and time-consuming
Solution Approach 1:
The system performs preliminary actions by pre-calculating and storing the optimal container numbers for various input load scenarios. This eliminates the need for time-consuming sequential scaling trials, as the appropriate container configuration is immediately available from the stored information when scaling is needed.
Solution Approach 2:
The system uses feedback by monitoring the current input load and comparing it against stored configurations to determine the appropriate container number. This feedback mechanism enables the service management device to make informed scaling decisions based on actual service conditions rather than following a fixed sequential scaling pattern.
Data Source
AI summary
Provided is a non-transitory computer-readable recording medium storing a service management program that causes a computer to execute a process, the process including acquiring a first input load indicating an amount of inputs received by a service at a first point in time, the service being implemented by containers, identifying first numbers of the containers corresponding to the first input load by referring to a storage unit that stores information where a second input load is associated with second numbers of the containers, the second input load indicating an amount of inputs received by the service when a response time of the service is reduced by increasing numbers of the containers to the second numbers of the containers in each of second points in time prior to the first point in time, and increasing the numbers of containers to the first numbers of the containers.