Workload Resource Profile Optimization for Time-Inflexible Data Center Tasks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods fail to optimize the utilization of resources for time-inflexible workloads in data centers and cloud storage systems, leading to resource waste due to inflexible workload requirements and subjective pricing, which are not applicable to time-inflexible workloads.
Innovation Solution
A method that calculates standard deviations for workload-consumed resource profiles, sorts them, and combines profiles that meet specific criteria to reduce resource usage, ensuring that combined profiles have lower maximum amounts and standard deviations, thereby optimizing resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If resources are allocated to meet peak requirements of time-inflexible workloads, then workload performance requirements are satisfied, but resource waste increases
Solution Approach 1:
The patent combines multiple time-inflexible workloads into workload groups based on compatibility analysis. By merging workloads with similar resource requirements and time constraints into groups, the system can allocate resources more efficiently at the group level rather than individually, reducing overall resource waste while still meeting each workload's performance requirements.
Solution Approach 2:
The patent changes the allocation parameter from individual workload-level resource allocation to group-level resource allocation. By analyzing compatibility parameters between workloads and allocating resources to compatible groups, the system optimizes resource utilization while ensuring time-inflexible workload requirements are met.
2Reliability
If more resources are allocated to ensure peak performance, then workload requirements are met, but cost of power consumption and maintenance increases
Solution Approach 1:
The patent merges multiple workloads into compatibility groups to consolidate resource allocation. By allocating resources to groups rather than individually to each workload, the system reduces redundant resource provisioning, thereby lowering power consumption and maintenance costs while ensuring all workload requirements are fulfilled.
Solution Approach 2:
The patent creates resource allocation groups that can serve multiple workloads simultaneously. Each resource allocation group is designed to handle multiple compatible workloads, making the resource allocation universal and multi-functional, which reduces the total resources needed compared to dedicated allocation for each workload.
3Productivity
If resources are dynamically deployed to catch up with workload changes, then resource optimization is achieved, but response speed to peak requirements decreases
Solution Approach 1:
The patent performs preliminary compatibility analysis and pre-establishes workload groups before peak demands occur. By analyzing workload compatibility in advance and pre-configuring resource allocation groups, the system is prepared to quickly allocate resources when peak requirements arise, achieving both optimization and fast response.
Solution Approach 2:
The patent implements dynamic workload group formation and resource allocation. The system can dynamically adjust which workloads are grouped together and reallocates resources based on current workload conditions, enabling both resource optimization and rapid response to changing demands.
4Adaptability or versatility
If subjective pricing methods are used for resource allocation, then flexible workload scheduling is achieved, but objectivity and consistency in resource allocation deteriorates
Solution Approach 1:
The patent enables workloads to self-categorize into compatibility groups based on their own characteristics and requirements. Each workload can identify its compatibility group through automated analysis, providing objective and consistent resource allocation without subjective pricing decisions, while still maintaining scheduling flexibility.
Solution Approach 2:
The patent implements feedback mechanisms that continuously analyze workload performance and resource utilization. This feedback loop provides objective data-driven insights for resource allocation decisions, replacing subjective pricing methods with measurable, consistent criteria while preserving scheduling adaptability.
Data Source
AI summary
Methods for optimizing utilization of workload-consumed resources for time-inflexible workloads are disclosed. By sorting workload-consumed resource profiles representing individual workloads in one system according to an order of standard deviation or descending volume, two workload-consumed resource profiles can be combined to check if combination criteria are fulfilled. If any combination satisfies the combination criteria, corresponding workloads can be combined to share the same resource from the system. Thus, optimizing utilization of the workload-consumed resource can be achieved.