Dynamic Workload Performance Goal Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for managing workload performance on mainframe computers, such as IBM System z, often result in sub-optimal performance due to static performance goal definitions, leading to undesirable outcomes like slower high-importance workloads and 'giving up' on underachieving service classes, despite features like service class definitions and Workload Manager (WLM) importance levels.
Innovation Solution
A system and method that dynamically assesses and adjusts service class performance goals based on achievement, criticality, and capacity shortages, using a management controller and agents to automatically change defined performance goals and notify users, optimizing workload performance across logical partitions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static performance goal definitions are used in service classes, then system configuration is simple and stable, but workload performance becomes sub-optimal and high-importance workloads slow down
Solution Approach 1:
The patent implements dynamic performance goal adjustment by continuously monitoring workload performance metrics and automatically modifying service class performance goals based on actual system conditions. This transforms the static performance goal definition into a dynamic system that adapts to changing workload patterns, thereby improving workload performance without requiring complex manual reconfiguration.
Solution Approach 2:
The system establishes a feedback loop where workload performance is continuously measured against defined goals, and the results feed back into automatic adjustments of performance parameters. This closed-loop control mechanism enables the system to self-optimize performance based on actual conditions, resolving the contradiction between simplicity and performance optimization.
2Productivity
If the Workload Manager gives up on underachieving service classes, then system stability is maintained, but performance optimization opportunities are lost
Solution Approach 1:
The patent enables the workload management system to automatically identify underachieving service classes and self-correct their performance goals without requiring external intervention or giving up on them. The system performs self-diagnosis and self-optimization by analyzing performance data and automatically adjusting parameters, thereby maintaining both automation and continuous performance improvement.
Solution Approach 2:
The system takes preliminary actions by proactively identifying service classes that are at risk of underachievement before performance degradation occurs. By monitoring trends and predicting potential performance issues, the system can preemptively adjust performance goals and allocate resources, preventing underachievement rather than reacting after the fact.
3Ease of operation
If manual performance goal adjustment is required, then system control is precise, but operational complexity and time consumption increase
Solution Approach 1:
The patent implements self-service performance optimization where the system automatically monitors, analyzes, and adjusts performance goals without requiring manual user intervention. This eliminates the time-consuming manual adjustment process while maintaining precise control through automated decision-making algorithms that base adjustments on real-time performance data.
Solution Approach 2:
The patent replaces the mechanical manual adjustment process with an automated electronic control system. Instead of requiring operators to manually analyze performance data and adjust parameters, the system uses automated monitoring and control mechanisms that continuously optimize performance, thereby eliminating time loss while maintaining operational precision.
Data Source
AI summary
In a system and method for managing mainframe computer usage, preferred values for service class defined performance goals are determined to optimize workload performance in service classes across a logical partition. A method for managing mainframe computer system usage can include receiving a performance optimization goal for workload performance in a service class, the service class having a defined performance goal. Achievement of the performance optimization goal is assessed, and a preferred value for the defined performance goal is determined based on assessing achievement of the performance optimization goal. Workload criticality can be taken into account, and automatic changes to the performance goal authorized.