Partitioned Computing Resource Volatility Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Partitioned computing systems face challenges in managing resources effectively due to unpredictable usage patterns, leading to potential underutilization or overutilization, which can impact performance and violate service level agreements (SLAs).
Innovation Solution
A method and system for managing resources in partitioned computing systems by computing resource usage volatility, determining the risk of resource saturation, and dynamically adjusting resource allocation through resource addition, removal, or sharing based on statistical analysis and user intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If resources are assigned to partitions based on SLAs, then service level requirements are met, but resource utilization becomes low and underutilization occurs
Solution Approach 1:
The patent implements dynamic resource allocation by continuously monitoring resource usage patterns and automatically adjusting resource assignments between partitions. The system transitions from static SLA-based allocation to dynamic allocation that responds to actual usage conditions, allowing resources to be reallocated in real-time to match demand while maintaining SLA compliance.
Solution Approach 2:
The system establishes a feedback loop where resource usage is continuously measured, volatility is calculated, and allocation decisions are made based on this feedback. The feedback mechanism compares actual usage against SLA requirements and automatically adjusts resource distribution to optimize utilization while ensuring service level compliance.
2Productivity
If less than required resources are assigned to partitions, then resource utilization risk increases and performance is affected, but assigning more resources ensures adequate performance
Solution Approach 1:
The system performs preliminary analysis of resource usage patterns and volatility calculations before making allocation decisions. By anticipating future resource needs based on historical data and statistical analysis, the system can proactively adjust resource allocation to prevent both over-utilization and under-utilization scenarios.
Solution Approach 2:
The patent changes the parameter of resource allocation from fixed SLA-based values to dynamic values based on measured usage and volatility. The system adjusts allocation parameters continuously based on actual performance data, allowing flexible adaptation to changing workloads while maintaining optimal resource utilization.
3Ease of operation
If resource allocation is fixed based on SLAs, then management is simple, but the system cannot adapt to changing usage patterns and volatility
Solution Approach 1:
The system implements self-service resource management where the partitioned computing system automatically monitors its own resource usage, calculates volatility, and adjusts allocation without external intervention. This automated self-adjustment maintains simplicity for users while providing sophisticated adaptability to changing usage patterns through machine-driven decision making.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method for managing resources in a partitioned computing system using determined risk of resource saturation is disclosed. In one example embodiment, the partitioned computing system includes one or more partitions. A volatility of resource usage for each partition is computed (102-108) based on computed resource usage gains/losses associated with each partition. A current resource usage of each partition is then determined (110). Further, a risk of resource saturation is determined (112) by comparing the computed volatility of resource usage with the determined current resource usage of each partition. The resources in the partitioned computing system are then managed (114) using the determined risk of resource saturation associated with each partition.