Software Defined Infrastructure Logical Resource Pools
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Prior art systems require full capacity of a physical server's resources to be configured for a single cloud solution, leading to under-utilization of available compute resources, as they do not support dedicating only a subset of resources to a cloud deployment or workload.
Innovation Solution
A software defined infrastructure (SDI) that allows a subset of a computer server's resources to be made available to cloud solutions or workloads, enabling the creation of logical resource pools across physical boundaries, which can be dynamically selected based on policies matching workload requirements with available resource features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full capacity of a physical server's resources is configured for a single cloud solution, then the server can be dedicated to that specific workload, but the available compute resources are under-utilized because they cannot be shared with other workloads
Solution Approach 1:
The patent segments physical server resources into multiple virtual server instances, each capable of being dedicated to different workloads. The resource manager divides compute resources (CPU, memory, storage) into allocable units that can be assigned to multiple virtual servers simultaneously, resolving the contradiction between dedication and utilization by allowing partial dedication to multiple workloads.
Solution Approach 2:
The resource manager implements multi-functionality by enabling physical server resources to serve multiple cloud solutions concurrently through virtualization. A single physical server can host multiple virtual servers, each dedicated to different workloads, making the physical resources universal and applicable to multiple purposes simultaneously, thus improving utilization while maintaining reliability through dedicated virtual instances.
2Productivity
If physical server resources are shared across multiple cloud solutions, then resource utilization improves, but the complexity of managing and allocating resources increases
Solution Approach 1:
The patent introduces a resource manager as an intermediary layer between physical server resources and multiple cloud solutions. This intermediary automatically handles resource allocation, monitoring, and management tasks, abstracting the complexity from users. The resource manager mediates between the physical resources and virtual servers, implementing policies for resource distribution without requiring manual intervention, thus improving utilization while managing complexity through automation.
Solution Approach 2:
The resource manager implements self-service capabilities by automatically allocating and managing resources based on predefined policies and workload requirements. The system autonomously monitors resource usage, performs dynamic allocation, and manages virtual server instances without requiring manual intervention, reducing management complexity while enabling efficient resource sharing across multiple cloud solutions.
3Ease of manufacture
If a physical server is dedicated to a single cloud solution, then resource allocation is simple, but the adaptability to support different workloads is reduced
Solution Approach 1:
The patent implements dynamic resource allocation where the resource manager can adjust resource distribution among virtual servers based on changing workload requirements. Physical server resources are not statically bound to a single cloud solution but can be dynamically reallocated between multiple virtual servers as needs change, providing both simplicity through automated management and adaptability through flexible resource distribution.
Data Source
AI summary
A software defined infrastructure (SDI) makes available a subset of a computer server's resources to a cloud solution or workload. Multiple subsets of resources can be combined in a SDI to provide a logical resource pool. This allows cloud administrators to create software defined infrastructures derived from the partial capacity of a collection of systems. The resources defined across the physical boundaries of a computer server can then be made available to host deployment of cloud workloads. The infrastructure resource pool can be selected upon deployment of a cloud workload.