Computational Resource Sub-allocation via Online Marketplace
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing infrastructure bills computing devices for reserved computational resources even if they are not used, leading to inefficiencies and unnecessary costs during non-peak workloads.
Innovation Solution
A method and system that allow computing devices to sub-allocate unused computational resources to other devices through an online marketplace, where resources are allocated based on actual usage and billed accordingly, enabling efficient utilization and cost recovery during peak workloads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If computing devices reserve predetermined computational resources for future peak workload, then reliability is improved, but loss of energy increases due to billing regardless of actual usage
Solution Approach 1:
The patent implements dynamic resource allocation where reserved computational resources can be flexibly sub-allocated to other computing devices during non-peak periods when the original device does not need them. The system monitors actual resource usage and dynamically adjusts allocation, allowing resources to transition from reserved to shared state based on real-time demand, thus eliminating waste while maintaining reliability during peak loads
Solution Approach 2:
The system changes the utilization parameter of reserved resources from static (always reserved) to dynamic (adjustable based on actual need). By implementing usage monitoring and conditional sub-allocation, the system transforms the state of resource reservation from a fixed commitment to a flexible arrangement where resources can be temporarily released for other uses and reclaimed when needed, resolving the contradiction between reliability and energy loss
2Productivity
If computing devices reserve computational resources in advance, then productivity is improved for peak workloads, but loss of time increases due to idle reserved resources
Solution Approach 1:
The patent ensures continuous useful action of computational resources by implementing a sub-allocation mechanism that prevents idle time. When reserved resources are not needed by the original computing device, they are automatically made available to other devices through the marketplace, ensuring that resources remain continuously productive. This eliminates the loss of time associated with idle reserved resources while maintaining the ability to provide full resources during peak workloads
3Reliability
If cloud infrastructure allocates computational resources based on SLA, then reliability is improved, but device complexity increases due to resource management overhead
Solution Approach 1:
The patent introduces an intermediary online marketplace platform that simplifies resource management complexity. The marketplace acts as a mediator between cloud infrastructure and computing devices, handling the complex tasks of resource tracking, sub-allocation, usage monitoring, and billing. This intermediary layer absorbs the management overhead, allowing individual devices to maintain simple SLA relationships with the cloud while the marketplace manages the complexity of dynamic resource sharing and coordination
Data Source
AI summary
The disclosed embodiments relate to systems and methods for method and systems for sub-allocating computational resources. A first computing device receives information associated with a first set of computational resources from a cloud infrastructure. The first set of computational resources has been allocated to the first computing device by the cloud infrastructure. A first set of parameters associated with a workflow received by the first computing device is determined. The first set of parameters is indicative of a need of the first set of computational resources by the first computing device. One or more computational resources from the first set of computational resources are sub-allocated based on the determined first set of parameters.