Job Allocation Support System for Server Resource Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulties in specifying an optimum server for job execution due to unclear job characteristics, leading to suboptimal cost performance in processing.
Innovation Solution
A job allocation support system that monitors resource usage across multiple servers, determines job characteristics based on resource use rates, and optimally reallocates jobs to servers with compatible characteristics and cost-effective resource configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If users specify allocation target servers manually, then ease of operation is maintained, but manufacturing precision deteriorates because users cannot accurately understand job characteristics
Solution Approach 1:
The system performs preliminary monitoring of resource use rates during job execution, accumulates execution history data, and determines job characteristics before making allocation decisions. This preliminary data collection and analysis enables accurate job-to-server matching without requiring users to manually analyze job characteristics.
Solution Approach 2:
The system introduces an intermediary allocation management device that automatically analyzes monitoring data, determines job characteristics, and recommends optimal server allocations. This intermediary handles the complex analysis work, freeing users from direct involvement while ensuring accurate allocation decisions based on objective data.
2Productivity
If job allocation is changed frequently to optimize cost performance, then productivity improves, but loss of time increases due to repeated reallocation operations
Solution Approach 1:
The system performs allocation optimization at periodic intervals based on accumulated monitoring data rather than continuously or frequently. By waiting for sufficient data to accumulate and analyze job characteristics, the system reduces the frequency of reallocation operations while still achieving cost optimization, thereby minimizing time loss from repeated changes.
Solution Approach 2:
The system uses feedback from monitoring resource use rates and execution history to make informed allocation decisions. By continuously monitoring and analyzing actual job performance data, the system optimizes allocations based on evidence rather than making frequent speculative changes, reducing unnecessary reallocation operations and associated time losses.
3Measurement precision
If comprehensive monitoring of resource use rates is performed, then measurement precision improves, but device complexity increases
Solution Approach 1:
The allocation management device performs multiple functions including monitoring resource use rates, determining job characteristics, analyzing execution history, and recommending allocations. By consolidating these functions into a single multi-functional system rather than separate specialized devices, the achievement of high measurement precision does not proportionally increase overall system complexity.
Data Source
AI summary
A system periodically or non-periodically performs allocation optimization support processing (that supports optimization of job allocation) based on monitoring result information (that indicates a result of monitoring of one or a plurality of jobs allocated to one or more servers). The allocation optimization support processing includes: determining, for each job based on the monitoring result information, a job characteristic compatible with resource use rates of various calculation resources in an allocation target server of the job; and selecting, when there is a job that matches a condition that change of the allocation target server is recommended, an allocation target server of the job after the change from among one or more servers having a server characteristic compatible with the job characteristic of the job based on at least one of an execution duration length and resource use rates of various calculation resources of the job, and a use cost of each of the one or more servers.