Data Processing System Iterative Object Redistribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-computing environments face challenges in efficiently redistributing objects across processing units, particularly in determining the optimal time for redistribution and minimizing downtime, while also ensuring even resource utilization and load balancing.
Innovation Solution
A data processing system that calculates the expected duration of redistribution and displays it to users, allowing informed decision-making on the start time, and offers modes for optimizing object distribution using either all or the minimum number of processing units, utilizing one-dimensional or two-dimensional distribution procedures based on object size and load properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If re-distribution of objects is performed to optimize resource utilization and load balancing, then system performance is improved, but downtime of processing units increases
Solution Approach 1:
The system performs a preliminary calculation of the expected duration required for re-distribution before actually executing the re-distribution. This allows administrators to plan and schedule re-distribution during periods of low system usage, thereby minimizing the impact of downtime on overall system productivity while still achieving optimized resource utilization and load balancing.
2Ease of operation
If re-distribution is performed during periods of low usage to minimize impact, then downtime impact is reduced, but system performance during re-distribution is worsened
Solution Approach 1:
The system provides feedback to administrators about the expected duration and impact of re-distribution operations. This information is used to make informed decisions about when to perform re-distribution, balancing the need to minimize operational impact with the need to maintain system performance. The feedback mechanism enables administrators to choose optimal timing that considers both operational convenience and performance requirements.
3Productivity
If all processing units are used for distribution, then resource utilization is optimized, but complexity of distribution increases
Solution Approach 1:
The patent segments the distribution process into distinct modes: a first mode that uses all available processing units for maximum resource utilization, and a second mode that uses only a minimum number of processing units for simpler distribution. This segmentation allows administrators to choose the appropriate level of complexity based on their specific needs, while still having access to optimized resource utilization when required.
Solution Approach 2:
The system dynamically adapts the distribution strategy based on the chosen mode. In the first mode, the system dynamically utilizes all processing units to optimize resource utilization. In the second mode, it dynamically selects only the minimum necessary processing units, thereby reducing distribution complexity. This dynamic approach allows the system to flexibly adjust between optimization and simplicity based on operational requirements.
Data Source
AI summary
A data processing system has a program module for determining a re-distribution of objects, such as database tables, over processing units of a cluster of processing units, such as in a blade computing environment. An estimate of the required duration for applying the re-distribution is calculated in order to facilitate an administrator's decision when to initiate the re-distribution procedure.