Hybrid Cloud Resource Allocator for Idle Private Compute
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Public clouds face challenges in providing adequate computing resources due to peak demands, and private clouds have limitations such as varying network connectivity and compute performance, making it difficult to efficiently utilize idle resources from private clouds for public cloud needs.
Innovation Solution
A system that opportunistically allocates computing resources by matching public cloud demands with temporarily available resources from private clouds, using a resource allocator and performance optimization analysis to ensure high performance and reliability, while also considering cost and dynamic availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If public clouds provide computing resources during peak demands, then service availability is improved, but resource adequacy deteriorates due to limited public cloud capacity
Solution Approach 1:
The patent combines public cloud and private cloud computing resources into a unified hybrid cloud system. The resource allocator integrates resources from both public cloud providers and private computing devices, allowing the system to leverage the scalability of public clouds while supplementing with private resources during peak demands, thereby resolving the contradiction between service availability and resource quantity.
Solution Approach 2:
The hybrid cloud system enables computing resources to serve multiple functions and multiple users simultaneously. Private computing devices that serve local users can also provide resources to public cloud services when idle, and public cloud resources can be dynamically allocated to different applications and users, maximizing resource utilization and addressing both availability and quantity constraints.
2Productivity
If private clouds utilize idle computing resources, then resource utilization is improved, but performance consistency deteriorates due to varying network connectivity and compute performance
Solution Approach 1:
The resource allocator implements continuous monitoring and feedback mechanisms to assess the performance and availability of private computing devices. Based on real-time feedback regarding network connectivity and compute performance, the system dynamically adjusts resource allocation decisions, directing workloads to private devices that currently meet performance thresholds while maintaining overall service consistency.
Solution Approach 2:
The system dynamically adapts resource allocation based on changing conditions. Private computing devices that temporarily meet performance requirements can be utilized, but the system remains flexible to switch between private and public cloud resources as performance conditions change, thereby maintaining performance consistency while improving overall resource utilization.
3Productivity
If the system opportunistically allocates resources from private clouds, then resource efficiency is improved, but system complexity increases due to resource matching and management overhead
Solution Approach 1:
The resource allocator serves as an intermediary component that manages the complexity of coordinating between public cloud providers and private computing devices. It handles resource discovery, capability assessment, matching, and allocation decisions, thereby encapsulating the complexity in a single manageable component while presenting a simplified interface to both public and private cloud participants.
Solution Approach 2:
Private computing devices autonomously expose their idle resource capabilities to the hybrid cloud system without requiring manual configuration or intervention. The resource allocator automatically discovers these resources, assesses their suitability, and incorporates them into the allocation pool, enabling efficient resource utilization while minimizing the operational complexity and management overhead for device owners.
Data Source
AI summary
A description of computing resource requirements for execution of an application associated with a publicly available service is obtained. Access to computing resources is opportunistically obtained from a computing entity that includes a private computing device that is external to, and separate from, the publicly available service. The computing resource requirements are intelligently matched to available computing resources of the computing entity with private computing resources that are temporarily available from a private computing device source. The intelligent matching is performed using an optimization analysis.